Journal of Ethnopharmacology 239 (2019) 111910

Contents lists available at ScienceDirect

Journal of Ethnopharmacology

journal homepage: www.elsevier.com/locate/jethpharm

Xian-Ling-Gu-Bao induced inflammatory stress rat liver injury: Inflammatory and oxidative stress playing important roles T

Wenxiao Wua,b, Ting Wangb, Bo Suna,b, Dong Liub, Zhi Linb, Yufa Miaob, Chao Wangb, ∗ ∗∗ Xingchao Gengb, ,BoLia,c, a Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, 100730, China b National Center for Safety Evaluation of Drugs, National Institute for Food and Drug Control, Key Laboratory of Beijing for Nonclinical Safety Evaluation of Drugs, A8 Hongda Middle Street, Beijing Economic-Technological Development Area, Beijing, 100176, China c National Institute for Food and Drug Control, 31 Hua Tuo Road, Daxing District, Beijing, 102629, China

ARTICLE INFO ABSTRACT

Keywords: Ethnopharmacological relevance: Xian-Ling-Gu-Bao (XLGB) Fufang is an herbal formula that has been used in Xian Ling Gu Bao clinical settings to treat osteoporosis, osteoarthritis, aseptic bone necrosis, and climacteric syndrome. Despite its Osteoarthritis uses, XLGB treatment has been linked to potential liver injury. To date, there is a lack of clear demonstration of Hepatotoxicity such toxicity in animal models. Inflammation factor Aim of the study: As animal models fail to reproduce the XLGB hepatotoxicity reported in humans, because Oxidative stress human hepatocytes are clearly more sensitive to XLGB, this study was designed to investigate a more reliable animal model of such toxicity. Materials and methods: We randomized rats into five groups, as follows: CON (control), XLGB, lipopolysaccharide (LPS), L-XLGB/LPS (XLGB, 0.125 g/kg; LPS, 0.1 mg/kg), and XLGB/LPS (XLGB, 1.25 g/kg; LPS, 0.1 mg/kg). These groups were treated with 0.5% sodium carboxymethyl cellulose (CMC-Na), XLGB suspension, normal saline, or LPS. The first administration of XLGB [0.125 g/kg or 1.25 g/kg, by mouth (PO)] or its solvent (0.5% CMC-Na) was delivered, and then food was removed. Twelve hours after the first administration of XLGB, rats received LPS [0.1 mg/kg, intravenously (IV)] or saline control. After 30 min, a second administration of XLGB (0.125 g/kg or 1.25 g/kg, PO) or solvent was administered. The rats were anesthetized at 12 h or 24 h following the second administration of XLGB. Liver function was evaluated by measuring liver weight, liver microscopy, serum biochemistry and plasma microRNA-122 (miR-122). The plasma levels of 27 cytokines were measured to evaluate inflammation. Moreover, the expression of cytochrome P450 2E1 (CYP2E1), nicotinic adenine dinu- cleotide phosphate (NADPH) oxidase and inducible nitric oxide synthase (iNOS) at protein levels were observed; immunofluorescence and immunohistochemistry were used to confirmed the hepatotoxicity of XLGB. Results: Hepatotoxicity in male rats with moderate inflammation induced by XLGB was indicated by liver his- topathology, serum biochemical analysis, serum miR-122 levels, and immunofluorescent assessments. We ob- served significant increases in liver weight and liver indexes in male rats with moderate inflammation in re- sponse to XLGB. Histopathological assessment further showed that extensive hepatocellular necrosis and inflammatory infiltration were evident in rats co-treated with XLGB/LPS. The levels of serum transaminases [alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT)], total bilirubin (TBIL) and triglyceride (TG), which are markers of liver function, were also significantly increased by XLGB/LPS treatment. Similarly, miR-122 was significantly elevated in XLGB/LPS treated rats relative to other groups. An immunofluorescent assessment showed extensive apoptosis in hepatocytes from these co-treated rats. What is more, XLGB can dose-dependently induce liver injury in male rats with moderate inflammation. Hepatic CYP2E1, neutrophil chemotactic factor (NCF-1), iNOS, and NOX-2 (an NADPH oxidase subunit) le- vels were increased in response to XLGB treatment, and staining for DMPO nitrone adducts further showed elevated oxidative stress level in XLGB/LPS-treated rats relative to the other experimental groups. Conclusion: LPS and XLGB co-treatment in rats led to marked hepatotoxicity. This toxicity was associated with disrupted lipid metabolism, extensive liver necrosis and inflammatory infiltration, apoptosis, and expression of oxidative stress-related proteins. These results demonstrate a valuable model for the study of iDILI in the context

∗ Corresponding author. ∗∗ Corresponding author. National Institute for Food and Drug Control, 31 Hua Tuo Road, Daxing District, Beijing, 102629, China. E-mail addresses: [email protected] (X. Geng), [email protected] (B. Li). https://doi.org/10.1016/j.jep.2019.111910 Received 25 December 2018; Received in revised form 31 March 2019; Accepted 23 April 2019 Available online 23 April 2019 0378-8741/ © 2019 Elsevier B.V. All rights reserved. W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

of XLGB treatment, and further provide insights into the potential mechanisms by which XLGB may induce hepatotoxicity in humans.

1. Introduction and death, systemic low levels of LPS can avoid causing acute in- flammation while still triggering signaling through appropriate detec- Osteoarthritis (OA) is a serious degenerative cartilage disease af- tion mechanisms (Miller et al., 2009). High-dose LPS is often used to fecting older adults (Musumeci et al., 2015), leading to potentially induce hepatotoxicity in animals, whereas low-dose LPS treatment is serious complications including bone marrow lesions (BMLs) and os- utilized to augment liver damage cause upon co-treatment with hepa- teophytosis, which, along with cystic lesions, are major causes of pain totoxic agents such as ethanol, carbon tetrachloride, halothane, cad- in those with OA (Kornaat et al., 2006). The risk factors for OA sus- mium, and allyl alcohol (Roth et al., 1997), or by treatment with drugs ceptibility include age, obesity, gender, and preexisting trauma. As the with idiosyncratic hepatotoxicity (Deng et al., 2009), such as ranitidine disease progresses, it leads to increasingly severe pain and joint dis- (Maddox et al., 2006), diclofenac (Deng et al., 2006), trovafloxacin orders, affecting the entirety of joints with further non-cartilage ar- (Shaw et al., 2009a)(Shaw et al., 2007), and sulindac (Zou et al., 2009). ticular pathologies, although the full extent of why and how OA pro- As LPS is widely used in animal models to assess hepatotoxicity, it is a gresses remains uncertain (Musumeci et al., 2013). valuable tool for assessing the potential toxicity of any drugs of interest. The adipocyte factors leptin, which is best understood in the context Xian-Ling-Gu-Bao (XLGB) Fufang is an herbal formulation approved of its ability to regulate metabolic processes, has also been found to be by the China Food and Drug Administration (CFDA; China, linked with OA development, as leptin levels are higher in OA patient Z20025337). As a herbal Fufang formulation, XLGB is composed of six serum and synovial fluid (Yan et al., 2018). Leptin is also linked with herbs: Epimedium brevicornu Maxim, foliage (70%); Dipsacus asper inflammatory responses, which can also drive OA development and Wall. ex C.B. Clarke, stem (10%); Salvia miltiorrhiza Bunge, root and progression (Scotece et al., 2014). For example, one study found that rhizome (5%); Anemarrhena asphodeloides Bunge, rhizome (5%); synovial fibroblasts exhibited dose-dependent IL-6 secretion when ex- Cullen corylifolium (L.) Medik, fructus (5%); and Rehmannia glutinosa posed to leptin (Yang et al., 2013), and interleukin (IL)-6 is in- (Gaertn.) DC, root and rhizome (5%). It has been used clinically to treat flammatory and prevalent in OA patinet serum and synovial fluid osteoporosis, osteoarthritis, aseptic bone necrosis, and climacteric (Kaneko et al., 2000). Toll-like receptors (TLRs) on articular chon- syndrome (Wu et al., 2009)(Tian et al., 2011)(Qiuqiong, 2010)(Lou drocytes are also linked with synovial fluid inflammation in OA patients et al., 2009)(Li and Hao, 2007)(He et al., 2008) over the past 20 years. (Schelbergen et al., 2012)(Kim et al., 2006), as such TLR signaling is Preclinical studies in ovariectomized (OVX) rats (Wu et al., 2017) and closely linked with OA pathogenesis (Scanzello et al., 2008). clinical trials (Li et al., 2018)(Zhang et al., 2007)(Zhu et al., 2012)have TLRs are a family of pattern recognition receptors (PRRs). TLR-4 is demonstrated the clinical efficacy of this agent, however on December able to specifically recognize bacterial lipopolysaccharide (LPS) and 8, 2016, the CFDA issued notification of a risk of liver damage caused trigger pro-inflammatory cytokine and chemokine production by cells by XLGB capsules. However, in Sprague-Dawley (SD) rats or OVX SD on which the receptor has been activated (Janssens and Beyaert, 2003) rats, XLGB has been reported to be safe up to 1800 mg/kg (Wang et al., (Kanzler et al., 2007). LPS is a large membrane-bound molecule pro- 2015)(Wu et al., 2017). In our hands we also failed to observe any liver duced by gram-negative bacteria. Given its pro-inflammatory potential, injury in 12-month old female rats following 5000 mg/kg XLGB treat- LPS has commonly been used in animal models to induce experimental ment for 90 days (data not shown). As animal models fail to reproduce arthritis via co-administration of LPS and collagen (Yoshino et al., the XLGB hepatotoxicity reported in humans, with human hepatocytes 2000)(Lorenz et al., 2013). As the large intestine and terminal ileum clearly being more sensitive to XLGB, there is a clear requirement for contain many gram negative bacteria, there is abundant LPS available more reliable animal models of such toxicity. in vivo, and intestinal LPS absorption (Fox et al., 1990) can induce low- Idiosyncratic drug-induced liver injury (iDILI) is often not effec- grade inflammation (Harte et al., 2010)(Brun et al., 2007)(Creely tively identified in the course of preclinical testing, and, as such, is a et al., 2007). While high-LPS internalization can lead to sepsis, shock, serious public health risk. Given the species-specificdifferences in

Fig. 1. The chromatograph of XLGB by HPLC.

2 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

XLGB toxicity, XLGB hepatotoxicity may be related to iDILI (QI and Evaluation of Drugs, Beijing, China approved all aspects of this study. REN, 2016). Based on the pathophysiology of OA and the standard approach of using low-dose LPS to improve detection of iDILI, we chose 2.4. Liver histopathology to test whether inducing a moderate degree of inflammation in rats would be sufficient to increase liver sensitivity to XLGB toxicity, al- 2.4.1. Liver weight lowing us to explore the mechanisms governing this toxicity. After euthanasia, livers were collected, weighed, observed, and photographed. A liver organ index was determined based on the weight 2. Materials and methods of the liver (g) divided by the body weight of the rat (g). Liver segments were then snap frozen with liquid nitrogen, while and other sections 2.1. Chemicals were fixed with 10% formalin to assess liver histopathology.

XLGB came from Guizhou Tongjitang Pharmaceutical, Co., Ltd. (Lot 2.4.2. Liver microscipy NO.1605016). The preparation was as follows (WS-10269 (ZD-0269)- Formalin-fixed liver sections were dehydrated and paraffin em- 2002): Dipsacus asper Wall. ex C.B. Clarke, stem (167 g), Salvia miltior- bedded, and a microtome was used to prepare 5-μm-thick sections. rhiza Bunge, root and rhizome (83 g), and Cullen corylifolium (L.) Medik, Sections were dewaxed, stained with hematoxylin and eosin (H&E), and fructus (83 g) were comminuted into fine powder; Epimedium brevicornu examined by a trained histopathologist in a blinded manner. Livers Maxim, foliage (1167 g), Anemarrhena asphodeloides Bunge, rhizome were scored based on hepatocellular lesion severity in three random (83 g), Rehmannia glutinosa (Gaertn.) DC, root and rhizome (83 g) were fields under 200×magnification. The level of necrosis and in- decocted with water for three times, 1 h/per time, then combined and flammatory infiltration were defined as follows: less than 1% of organ concentrated until the density reached 1.35–1.38 g/mL (30 °C). The fine affected, scored as 0; minimal, the lesions easily recognizable, but to a power added to the extractum, made into pellets, and dried. The final lesser extent, from 1% to 25% of organ affected, scored as 1; mild, from weight was 500 g. 25% to 50% of organ affected, scored as 2; moderate, from 50% to 75% To determine the chemical constituents, XLGB was analyzed with of organ affected, scored as 3; marked, from 75% to 100% of organ HPLC [performed as Drug Standard of State Food and Drug affected, scored as 4 (Drury, 1980)(Hardisty and Eustis, 1990). The Administration (WS-10269 (ZD-0269)2002-2011Z) and Chinese rate of necrosis or inflammatory infiltration was determined based on Pharmacopoeia (2015 edition)]. The content of Epimedin C was 3.7 the number of rats with lesions with more than 1% of organ affected mg/0.5 g and Icariin was 0.7 mg/0.5 g (Fig. 1). divided by the total number of rats in each group. The LPS (Lot No. L-2880, derived from Escherichia coli O55:B5, purified by phenol extraction) came from Sigma-Aldrich (MO, United 2.5. Blood analysis States). The goat anti-NCF1 (ab795), rabbit anti-iNOS (ab15323), rabbit anti-NOX2/gp91phox (ab31092), mouse anti-3-Nitrotyrosine 2.5.1. Serum biochemistry [39B6] (ab61392), mouse anti-CD68 [KP1] (ab955) and rabbit anti- Blood was collected without anticoagulant, and allowed to coagu- alpha smooth muscle actin (а-SMA, ab5694) came from Abcam late for 30 min, after which it was spun for 10 min at 3000 rpm to (Cambridge, UK). isolate serum. The serum was then used to measure serum alanine transaminase (ALT), aspartate transaminase (AST), gamma-glutamyl 2.2. Animals transferase (GGT), total bilirubin (TBIL), and triglyceride (TG) levels with a HITAC7180A automatic analyzer (Hitachi, Japan). ALT and AST We purchased male SD rats (180–220 g; 6-weeks old) from Vital served to assess liver damage, while TBIL and GGT were indicators of River Laboratories (Beijing, China). The rats were house under standard liver secretory function, and TG and MDA were representative of lipid conditions in a specific pathogen free (SPF) facility, with three rats per metabolism. cage. Before beginning the study, animals were given 10 days to ac- climatize. Rats received ad libitum water and food access, and were 2.5.2. Plasma microRNA-122 (miR-122) identified using ear tags. Total plasma miRNA was isolated via thr miRNeasy Serum/Plasma Kit (QIAGEN, Lot No. 217184) according to provided directions. For 2.3. Study design reverse transcription, RNA, ID3EAL RT Primer 1-plexes (20×) (MiRXES, Lot No.1103113), ID3EAL Spike-in control (MiRXES, Lot We randomized rats into five groups: CON (control), XLGB (1.25 g/ No.1102122), ID3EAL miRNA RT Buffer (4×) (MiRXES, Lot kg/per time, twice daily), LPS, L-XLGB/LPS (XLGB, 0.125 g/kg/per No.1103103), and nuclease free water were mixed in 15 μL. The reac- time, twice daily; LPS, 0.1 mg/kg) and XLGB/LPS (XLGB, 1.25 g/kg/ tion proceeded at 42 °C for 30 min followed by heat-inactivation at per time, twice daily; LPS, 0.1 mg/kg). These groups were treated with 95 °C for 5 min. To amplify and detect reverse transcribed miRNAs, 5 μL 0.5% CMC-Na, XLGB suspension, normal saline, or LPS. XLGB doses and of cDNA (diluted 1:10) was used for a quantitative polymerase chain times were selected based on the dose and time used in human clinical reaction (qPCR) reaction in a 15 μL volume using ID3EAL miRNA qPCR trials (1500 mg/60 kg/per time, twice daily). Based on rat-human body Master Mix (2×) and ID3EAL miRNA qPCR assays (10×) (MiRXES, Lot surface area conversion standards, 0.125 g/kg/per time is equal to the No.1104103). The thermocycler conditions were as follows: 10 min of dose used in humans. Rats received two administrations of XLGB. The denaturation at 95 °C, 5 min at 40 °C, followed by 40 cycles of 10 s at first administration of XLGB [0.125 g/kg or 1.25 g/kg, 20 mL/kg, by 95 °C and 30 s at 60 °C. mouth (PO)] or its vehicle (0.5% CMC-Na, 20 mL/kg, PO) was deliv- ered, and then food was removed. 12 h after the first administration of 2.5.3. Cytokine measurements XLGB, rats received LPS [0.1 mg/kg, 2 mL/kg, intravenously (IV)] or Plasma levels of 27 cytokines including Eotaxin/CCL11, granulo- saline control. After 30 min, a second administration of XLGB (0.125 g/ cyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony kg or 1.25 g/kg, 20 mL/kg, PO) or solvent was administered (Zou et al., stimulating factor (GM-CSF), interferon-gamma (IFN-γ), interleukin- 2009). Rats were anesthetized with pentobarbital sodium and eu- 1alpha (IL-1 α), interleukin-1beta (IL-1β), IL-2, IL-4, IL-5, IL-6, IL-10, IL- thanized at 12 h or 24 hous following the second administration of 12p70, IL-13, IL-17A/CTLA8, IL-18, endothelial growth factor (EGF), XLGB. Blood was then collected either with ethylenediamine tetraacetic IP-10/CXCL10, leptin, MIP-1α/CCL3, RANTES/CCL5, tumor necrosis acid (EDTA) or without anticoagulants for analysis. The Institutional factor-alpha (TNF-α), vascular endothelial growth factor type A (VEGF- Animal Care and Use Committee of the National Center for Safety A), KC/GRO/CINC-1/CXCL1, Fractalkine/CX3CL1, LIX, MIP-2/CXCL2,

3 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910 and MCP-1/CCL2 were determined simultaneously using a magnetic 3. Results bead panel kit (Merck Millipore; Lot No. RECYTMAG-65K-27) based on the provided instructions. A Bio-Plex 200 System (Bio-Rad, USA) was 3.1. XLGB/LPS-induced liver histopathological damage used for data collection, and Bio-Plex Manager 5.0 was used for data analysis via five-parameter logistic regression curve fitting. For data Following administration of XLGB and/or LPS, rats were euthanized interpretation, any cytokines below the limit of detection for the assay and livers were collected and examined. As shown in Table 1, liver were assigned the minimum value, whereas those greater than 3 stan- weights and liver index values were elevated significantly at 12 h after dard deviations above the mean were omitted from downstream cal- the second XLGB administration. culations. Macroscopic observations showed white spots on the surface of the livers in the L-XLGB/LPS and XLGB/LPS groups. The density of the white spots is dependent to the dose of XLGB. (Fig. 1A). A histo- 2.6. Western blotting pathological assessment determined that whereas pathology in the CON group was normal, the XLGB and LPS groups showed slight evidence of Snap frozen liver tissue was homogenized and then lysed on ice for inflammation. Moreover the L-XLGB/LPS and XLGB/LPS co-treated ff 30 min using a tissue lysis solution (1:9 dilution of liver: lysis bu er). group showed evidence of severe widespread tissue damage, and both Samples were spun (14,000 g, 25 min, 4 °C) and supernatants were the proportion and incidence rate in the XLGB/LPS group are higher collected. Supernatant protein levels were assessed via BCA assay, and than those in the L-XLGB/LPS group (Fig. 2B, C, 2D, 2E, 2F). protein samples were them separated on SDS-PAGE gels via electro- phoresis. Next, samples were transferred for PVDF membranes, which 3.2. XLGB/LPS-induced liver dysfunction were blocked via 5% nonfat dry milk powder for 2 h and then probed overnight using mouse anti-GAPDH, rabbit anti-CYP2E1, goat anti- The liver is essential for the metabolism of vital macromolecules NCF1/p47phox, rabbit anti-NOX2/gp91phox, or rabbit anti-iNOS. Blots including proteins, carbohydrates, amino acids, and lipids, with the were probed using an appropriate horseradish peroxidase (HRP)-con- liver transaminases TBIL and TG serving as classical biomarkers of liver jugated secondary antibody. injury (Johnston, 1999). Serum ALT and AST are similarly markers of liver cell damage. We found that serum ALT, AST, GGT, TBIL, and TG fi 2.7. Immunofluorescence analysis levels were signi cantly elevated in rats in the L-XLGB/LPS and XLGB/ LPS groups relative to the other groups (Fig. 3A–E). Serum GGT and fi Liver sections were deparaffinized and stained with a standard im- TBIL were signi cantly elevated following XLGB/LPS co-treatment, munostaining protocol using primary anti-DMPO nitrone adduct at the suggesting that XLGB/LPS co-treatment disrupted bile acid secretion fi recommended dilution. Sections were mounted with ProLong Gold and excretion in liver. Similarly, the signi cant elevation in TG levels in Antifade Reagent and were stained with DAPI prior to examination the L-XLGB/LPS and XLGB/LPS groups suggested a disruption in lipid under a 20x objective. metabolism. MiR-122 is a highly conserved and abundant liver-specific miRNA, the plasma levels of which increase following liver injury in both ro- 2.8. Immunohistochemistry (IHC) analysis dents and humans (Starkey Lewis et al., 2011)(Wang et al., 2009). As shown in Fig. 4, LPS treatment led to elevated miR-122 levels as Tissue sections were prepared as described, and stained using pri- compared to CON or XLGB treatment, whereas XLGB/LPS co-treatment mary anti-CD68 and anti-а-SMA, followed by appropriate secondary led to even more significant plasma miR-122 elevation. antibodies. A 20× objective was used for imaging. 3.3. Kupffer and hepatic stellate cell activation following XLGB/LPS challenge 2.9. Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling (TUNEL) staining We next assessed how XLGB/LPS co-treatment affected the Kupffer and hepatic stellate cell activation by assessing the expression of CD68 Tissues sections were prepared as described and stained to identify and а-SMA, which are markers of Kupffer and hepatic stellate cell ac- apoptotic hepatocytes with a TUNEL Detection Kit (GeneTex, tivation, respectively. We observed significant increases in both of these GTX85585) based on provided instructions. A 20×objective was used markers in livers of rats co-treated with XLGB/LPS compared with those for imaging. of rats treated CON, XLGB, or LPS alone (Fig. 5).

2.10. Statistical analysis 3.4. Apoptosis induced by XLGB/LPS hepatotoxicity

All data are given as means ± standard error of the mean (SEM). In addition to necrosis, apoptosis is a principal mode of cellular Student's t-tests were used for comparing two groups, and one-way death in the context of liver injury; thus we used a TUNEL assay to analyses of variance (ANOVAs) were used to compare treated and un- assess apoptosis in the context of XLGB/LPS-induced liver damage. To treated groups. GraphPad Prism 6 (GraphPad Software, Inc., San Diego, further confirm apoptosis, we also assessed the levels of hepatic NF-κB CA, United States) was used for all analyses, with P < 0.05 as the and Caspase-3. Relative to the CON, XLGB, and LPS-only groups, the threshold of statistical significance. XLGB/LPS co-treated liver samples showed elevated NF-κB and

Table 1 Effects of XLGB/LPS on liver weight (g) and liver index (%).

Parameter CON XLGB LPS L-XLGB/LPS XLGB/LPS

Liver weight(g) 8.49 ± 0.12 8.38 ± 0.19 9.20 ± 0.16** 9.66 ± 0.20***,# 9.46 ± 0.17*** Liver index (%) 2.90 ± 0.04 2.90 ± 0.05 3.30 ± 0.06*** 3.40 ± 0.07*** 3.30 ± 0.06***

**, P < 0.01 (vs. control), ***, P < 0.001 (vs. control), ##, P < 0.05 (vs. LPS).

4 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Fig. 2. Liver histopathology following XLGB/LPS co-treatment. (A) Macropathology of livers (B) Liver sections were dewaxed and stained using H&E. Inflammatory infiltration (yellow arrows), necrosis (red arrows) are indicated. (C) The incidence rate of necrosis. (D) The necrosis score. (E) The incidence rate of inflammatory infiltration. (F) The inflammatory infiltration score. Data are means ± SEM, n = 10 for each bar. ##, P < 0.01 (vs. LPS). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Caspase-3 protein levels (Fig. 6), suggesting that apoptosis is a sig- stellate cells, were significantly elevated following XLGB/LPS co- nificant mode of cell death in the liver following XLGB/LPS treatment. treatment as compared with CON, XLGB, and LPS treatment.

3.5. Effects of XLGB on LPS-induced cytokine production 3.6. Oxidative stress is important for XLGB/LPS-induced acute liver injury

LPS is known to induce the secretion of many different cytokines, Oxidative stress is known to be important for hepatic injury, with with different kinetics for each cytokine (Biesmans et al., 2016). Certain important roles for cytochrome P450, NADPH oxidase, and iNOS in cytokines are thought to be linked to drug/liver interactions; thus, we iDILI pathology and drug metabolism (Aubert et al., 2011). Because therefore assessed plasma levels of 27 different cytokines following 12 h leptin can induce expression of NADPH oxidase and iNOS, this could XLGB/LPS co-treatment. The concentrations of leptin, IL-6, and MCP-1 induce oxidative stress, leading to the production of peroxynitrite, a (Fig. 7A–C; other cytokines in Fig. SI2), all of which regulate hepatic strong physiological oxidant.

5 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Fig. 3. Liver functions in response to XLGB/LPS co-treatment (A) ALT. (B) AST. and (C) GGT levels in serum were measured as markers of liver toxicity. (D) The level of serum TG. (E) The level of serum TBIL. (F) The level of plasma MDA, n = 4 per bar. Data are means ± SEM, n = 10 for each bar. *, P < 0.05, **, P < 0.01, ***, P < 0.001, (vs. control); #, P < 0.05, ##, P < 0.01 (vs. LPS).

model in reproducing XLGB hepatotoxicity has limited the studies of underlying hepatotoxic mechanisms. To better sensitize rat hepatocytes to XLGB hepatotoxicity, we administered low-dose LPS to rats receiving XLGB. In so doing, we were able to confirm that XLGB can induce he- patotoxicity in male rats by liver histopathology, serum biochemical analyses, serum miR-122 levels, and immunofluorescent assessments. Elevated levels of transaminase, total bilirubin, and liver cell injury have been found in patients with severe XLGB-induced hepatotoxicity. Consistent with this, we observed significant increases in liver weight and liver indexes in male rats with moderate inflammation in response to XLGB. Histopathological assessment further showed that extensive hepatocellular necrosis and inflammatory infiltration were evident in rats with XLGB/LPS co-treatment. The levels of serum transaminases (ALT, AST, GGT), TBIL, and TG, which are markers of liver function, Fig. 4. The level of plasma miR-122. Data are means ± SEM, n = 10 for each were also significantly increased by XLGB/LPS treatment. Similarly, bar. *, P < 0.05, **, P < 0.01, ***, P < 0.001, (vs. control); #, P < 0.05, miR-122, a sensitive circulating biomarker of liver injury (Sharapova ##, P < 0.01 (vs. LPS). et al., 2016)(Starckx et al., 2013), was significantly elevated in XLGB/ LPS treated rats relative to other groups. Similarly, an immuno- We found that hepatic CYP2E1, NCF-1, iNOS, and NOX-2 (an fluorescent assessment showed extensive apoptosis in hepatocytes from NADPH oxidase subunit) levels were increased in response to XLGB these co-treated rats, thus indicating that both necrotic and apoptotic treatment (Fig. 8A). Staining for DMPO nitrone adducts further showed cell death are induced by this drug co-treatment combination. More- elevated levels in XLGB/LPS co-treated rats relative to the other ex- over, XLGB can dose-dependently induce liver injury in male rats with perimental groups (Fig. 8E). moderate inflammation. Kupffer cells are specialized liver macrophages that line the walls of the sinusoids, where they mediate early liver injury responses. These 4. Discussion Kupffer cells are capable of internalizing endotoxin, activating the transcription and secretion of pro-inflammatory cytokines and super- XLGB has been used in clinical trials successfully for over 20 years; oxides, which can in turn activate stellate cells (Wheeler, 2003). In the however, recent concerns over its potential hepatotoxicity have been context of liver fibrosis, stellate cells are the primary cells responsible raised and require further examination. The lack of an available animal

6 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Fig. 5. Kupffer and Hepatic stellate cell activation. (A) Normalized CD68 expression (normalized to CON). (B) Liver CD68 was determined via western blot, with GAPDH as a loading control. (C) Normalized a-SMA expression (normalized to CON). (D) Liver a-SMA via western blot, with GAPDH as a loading control. (E–F) Representative liver section images, with nuclei in blue and positive staining shown in brown for CD68 (E) and а-SMA (F). **, P < 0.01, vs. control, ##, P < 0.01, vs. LPS. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) for driving scar formation and consequent liver damage. These stellate Oxidative stress is a major contributor to the damage that arises in cells can also alter their metabolism to induce or drive liver injury response to chemicals and other hepatotoxic agents. While XLGB/LPS (Chen et al., 2012), and hepatocyte necrosis further influences this co-treatment induced acute liver damage, it has the potential to drive process (Das et al., 2013). fibrogenesis in rats. The hepatic microsomal cytochrome P450 All cytokine levels in serum increase significantly after administra- (CYP450) system is important for reactive oxygen production, and we tion of LPS. Various cytokines peak at different points, such as IL-6, found that in rats treated with XLGB, levels of CYP2E1, 1A2, 3A2, 3A1 TNF-α, and MCP-1 at 2 h after administration, and IFN-γ, and, leptin at and 2C6 protein were increased (data not shown). CYP2E1 is thought to 6 h later. Most cytokines returned to baseline 24 h later, whereas IL-6 contribute to increased liver sensitivity to hepatotoxic drugs, po- and MCP-1 were still at a high level (Biesmans et al., 2016)(Biesmans tentiating LPS-induced oxidative stress in the liver (Cederbaum et al., et al., 2013). Leptin can bind to Kupffer cells’ functional receptor in the 2012). CYP2E1 is also important for triglyceride accumulation, hepatic liver and induce the release of other cytokines from Kupffer cells (De steatosis, and apoptotic induction (Wu et al., 2010)(Abdelmegeed Minicis et al., 2008). The increased release of MCP-1 is one of the ac- et al., 2012)(Cederbaum, 2010). XLGB-induced CYP2E1 over- tivation markers of Kupffer cells (Chatterjee et al., 2013). Leptin also expression thus may enhance and exacerbate liver toxicity. We verified makes a contribution to the formation of free radicals in various cells, the potential role of CYP2E1 in XLGB/LPS-induced hepatotoxicity including hepatic stellate cells (De Minicis et al., 2008)(Cao et al., through the pre-administration of diallyl sulfide (DAS), a selective in- 2007). IL-6, MCP-1, and leptin can also activated hepatic stellate cells hibitor of CYP2E1, to rats, showing that CYP2E1 is an important (Aleffi et al., 2005)(Honda et al., 2002)(Tsuchida and Friedman, mediator of XLGB/LPS-induced hepatic toxicity (Fig. SI). 2017), and the activated hepatic stellate cells play an important role in Production of superoxide and nitric oxide by NADPH oxidase and the development of liver fibrosis (Hernandez-Gea and Friedman, 2011). iNOS can result in the production of the strong physiological oxidant XLGB co-treatment with LPS can increase the levels of IL-6, leptin, and peroxynitrite. Oxidative stress induced by NADPH oxidase and iNOS MCP-1 to go even higher and for a longer time, which magnifies the has been found to be important for liver damage, especially via driving liver injury. The hepatic steatosis, necrosis and apoptosic also provide Kupffer cell activation and fibrogenesis (De Minicis et al., 2008)(Paik positive feedback about the elevation of these cytokines. et al., 2014)(Chatterjee and Das, 2015). NADPH oxidase-mediated

7 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Fig. 6. Hepatocellular apoptosis following XLGB/LPS co-treatment. (A) NF-κB expression, normalized against CON. (B) Liver NF-κB as determined via western blot, with GAPDH as a loading control. (C) Caspase-3 expression, normalized against CON. (D) Liver Caspase-3 as determined via western blot, with GAPDH as a loading control (E) Hepacellular apoptosis was evaluated by TUNEL assay (original magnification × 200). peroxynitrite production facilitates the recruitment of TLR4 into he- 2016), and these factors can induce liver injury, particularly in the patic lipid rafts, thereby enhancing inflammation (Das et al., 2015) and context of drug interactions (Chatterjee et al., 2013)(Lu et al., 2013) promoting liver injury. Cytokines such as TNF-a, IL-6, MCP-1, and (Poulsen and Olivero-Verbel, 2014)(Shaw et al., 2009b)(Fredriksson leptin are upregulated after TLR4 activation by LPS (Biesmans et al., et al., 2011). We also found that NCF-1, NOX-2 and iNOS protein

Fig. 7. Cytokines production in response to LPS and XLGB. (A) Plasma leptin levels. (B) Plasma IL-6 levels. (C) Plasma MCP-1 levels. Data are shown as means ± SEM, n = 10 per bar. *, P < 0.05, **, P < 0.01, ***, P < 0.001 (vs. control); #, P < 0.05, ##, P < 0.01 (vs. LPS).

8 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Fig. 8. Oxidative stress in response to XLGB. (A) Immunoblotting analysis of NCF-1 following 24 h XLGB/LPS treatment. (B) Immunoblotting analysis of 3-NT following 12 h XLGB/LPS treatment. (C) Immunoblotting for CYP2E1, NCF-1, NOX-2, and iNOS following 12 h XLGB treatment. (D) NCF-1 and 3-NT levels in liver extracts were determined by western blotting, with GAPDH as a loading control. (E) Liver sections stained for DMPO-nitrone adducts. expression was induced by XLGB. MCP-1 plays an important role in one of the first-line treatments for osteoporosis and osteoarthritis. liver inflammation, and its release is considered a hallmark of Kupffer Given to the complexity of pathologic processes of osteoporosis and cell activation. CD68, another marker of Kupffer cell activation, was osteoarthritis, the predisposing factors need to be checked with the also significantly increased in the XLGB/LPS group compared with patient and the pathological examination (such as hematology, im- other groups. WB and IHC analysis further showed that α-SMA levels munoglobulin test) before the treatment. Once the patient has a history were elevated, indicating hepatic stellate cell activation. We also de- of allergy, or chronic liver and immune system diseases, cautions tected fibrogenesis in the XLGB/LPS group. should be exercised to avoid the liver injury induced by XLGB.In con- There are two forms of osteoporosis: Primary osteoporosis and clusion, we found that while individually LPS and XLGB were asso- secondary osteoporosis. The Primary osteoporosis is related to estrogen ciated with only minor hepatotoxicity, co-treatment of rats with these deficiency and natural aging and secondary osteoporosis is due to dis- two compounds led to clear marked hepatotoxicity. This toxicity was ease pathology (Collins et al., 2017; Lewiecki, 2008; Raisz, 2005; Tella associated with disrupted lipid metabolism, extensive liver necrosis and and Gallagher, 2014), such as diabetes. Gut microbiota dysregulation, inflammatory infiltration, apoptosis, and expression of oxidative stress- IL-6 (Quach and Britton, 2017) and the increases of ROS(Garrett et al., related proteins including CYP2E1, NADPH oxidase and iNOS. These 1990) and NADPH oxidases (Thrailkill et al., 2005) are correlated with results thus demonstrate a valuable model for the study of iDILI in the increased inflammatory response and bone resorption. For OA, both context of XLGB treatment, and further provide insights into the po- adipocyte factor leptin and IL-6 are linked with OA development, as tential mechanisms by which XLGB may induce hepatotoxicity in hu- leptin and IL-6 levels are higher in OA patient serum and synovial fluid mans. Therefore, when XLGB is used in combination therapy, especially (Yan et al., 2018)(Kaneko et al., 2000). As all known, Leptin and IL-6 is with an inducer of CYP2E1 (e.g., acetaminophen) or oxygenant, or the also linked with inflammatory responses (Scotece et al., 2014). XLGB is patients with a history of allergy, or chronic liver and immune system

9 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910 diseases, cautions should be exercised. In addition, patients need to Moylan, C.A., Garibaldi, F., Premont, R., Suliman, H.B., Piantadosi, C.A., Diehl, A.M., avoid drinks containing alcohol (inducer of CYP2E1) and fatty food to 2012. Hedgehog controls hepatic stellate cell fate by regulating metabolism. Gastroenterology 143, 1319–1329. e11. https://doi.org/10.1053/j.gastro.2012.07. reduce the risk of XLGB-induced liver injury. 115. Collins, F.L., Rios-Arce, N.D., Schepper, J.D., Parameswaran, N., McCabe, L.R., 2017. The Conflicts of interest potential of probiotics as a therapy for osteoporosis. Microbiol. Spectr. 5. https://doi. org/10.1128/microbiolspec.BAD-0015-2016. Creely, S.J., McTernan, P.G., Kusminski, C.M., Fisher, f, M., Da Silva, N.F., Khanolkar, M., The authors declare that they have no conflicts of interest. Evans, M., Harte, A.L., Kumar, S., 2007. Lipopolysaccharide activates an innate im- mune system response in human adipose tissue in obesity and type 2 diabetes. Am. J. – Author contribution Physiol. Endocrinol. Metab. 292, E740 E747. https://doi.org/10.1152/ajpendo. 00302.2006. Das, S., Alhasson, F., Dattaroy, D., Pourhoseini, S., Seth, R.K., Nagarkatti, M., Nagarkatti, Wenxiao Wu, Xingchao Geng and Bo Li-Investigation, Formal ana- P.S., Michelotti, G.A., Diehl, A.M., Kalyanaraman, B., Chatterjee, S., 2015. NADPH fl lysis Writing editing (original draft); Wenxiao Wu, Ting Wang, Dong oxidase-derived peroxynitrite drives in ammation in mice and human nonalcoholic steatohepatitis via TLR4-lipid raft recruitment. Am. J. Pathol. 185, 1944–1957. Liu, Bo Sun and Chao Wang-dissecting the animals; Wenxiao wu-ad- https://doi.org/10.1016/j.ajpath.2015.03.024. ministration, western blot, RT-PCR, ICH, IF; Zhi Lin-Pathological Das, S., Seth, R.K., Kumar, A., Kadiiska, M.B., Michelotti, G., Diehl, A.M., Chatterjee, S., reading; Yufa Miao-serum biochemistry. 2013. Purinergic receptor X7 is a key modulator of metabolic oxidative stress- mediated autophagy and inflammation in experimental nonalcoholic steatohepatitis. Am. J. Physiol. Gastrointest. Liver Physiol. 305, G950–G963. https://doi.org/10. Acknowledgement 1152/ajpgi.00235.2013. De Minicis, S., Seki, E., Oesterreicher, C., Schnabl, B., Schwabe, R.F., Brenner, D.A., 2008. Reduced nicotinamide adenine dinucleotide phosphate oxidase mediates fibrotic and We would like to thank Guizhou Tongjitang Pharmaceutical for inflammatory effects of leptin on hepatic stellate cells. Hepatology 48, 2016–2026. their supporting. We would like to thank Lin Zhang, Dongsheng Pan, https://doi.org/10.1002/hep.22560. and Yanwei Yang for excellent technical assistant in animal and pa- Deng, X., Luyendyk, J.P., Ganey, P.E., Roth, R.A., 2009. Inflammatory stress and idio- syncratic hepatotoxicity: hints from animal models. Pharmacol. Rev. 61, 262–282. thological studies. This work was supported by National Major https://doi.org/10.1124/pr.109.001727. Scientific and Technological Special Project for “Significant New Drugs Deng, X., Stachlewitz, R.F., Liguori, M.J., Blomme, E.A., Waring, J.F., Luyendyk, J.P., Development” (No. 2015ZX09501004-002) from the Ministry of Maddox, J.F., Ganey, P.E., Roth, R.A., 2006. Modest inflammation enhances diclo- Science and Technology of the People's Republic of China. fenac hepatotoxicity in rats: role of neutrophils and bacterial translocation. J. Pharmacol. Exp. Ther. 319, 1191–1199. https://doi.org/10.1124/jpet.106.110247. Drury, R., 1980. Theory and Practice of Histotechnology. Appendix A. Supplementary data Fox, E.S., Thomas, P., Broitman, S.A., 1990. Hepatic mechanisms for clearance and de- toxification of bacterial endotoxins. J. Nutr. Biochem. 1, 620–628. Fredriksson, L., Herpers, B., Benedetti, G., Matadin, Q., Puigvert, J.C., de Bont, H., Supplementary data to this article can be found online at https:// Dragovic, S., Vermeulen, N.P., Commandeur, J.N., Danen, E., de Graauw, M., van de doi.org/10.1016/j.jep.2019.111910. Water, B., 2011. Diclofenac inhibits tumor necrosis factor-alpha-induced nuclear factor-kappaB activation causing synergistic hepatocyte apoptosis. Hepatology 53, 2027–2041. https://doi.org/10.1002/hep.24314. References Garrett, I.R., Boyce, B.F., Oreffo, R.O., Bonewald, L., Poser, J., Mundy, G.R., 1990. Oxygen-derived free radicals stimulate osteoclastic bone resorption in rodent bone in – Abdelmegeed, M.A., Banerjee, A., Yoo, S.H., Jang, S., Gonzalez, F.J., Song, B.J., 2012. vitro and in vivo. J. Clin. Investig. 85, 632 639. https://doi.org/10.1172/JCI114485. Critical role of cytochrome P450 2E1 (CYP2E1) in the development of high fat-in- Hardisty, J.F., Eustis, S.L., 1990. Toxicologic pathology: a critical stage in study inter- duced non-alcoholic steatohepatitis. J. Hepatol. 57, 860–866. https://doi.org/10. pretation. In: Progress in Predictive Toxicology. 1016/j.jhep.2012.05.019. Harte, A.L., da Silva, N.F., Creely, S.J., McGee, K.C., Billyard, T., Youssef-Elabd, E.M., Aleffi, S., Petrai, I., Bertolani, C., Parola, M., Colombatto, S., Novo, E., Vizzutti, F., Tripathi, G., Ashour, E., Abdalla, M.S., Sharada, H.M., Amin, A.I., Burt, A.D., Kumar, Anania, F.A., Milani, S., Rombouts, K., Laffi, G., Pinzani, M., Marra, F., 2005. S., Day, C.P., McTernan, P.G., 2010. Elevated endotoxin levels in non-alcoholic fatty fl Upregulation of proinflammatory and proangiogenic cytokines by leptin in human liver disease. J. In amm. 7, 15. https://doi.org/10.1186/1476-9255-7-15. ff hepatic stellate cells. Hepatology 42, 1339–1348. https://doi.org/10.1002/hep. He, Y.C., Liao, K.X., Huang, Y.R., Jiang, T.B., 2008. Curative e ect of xianlinggubao 20965. capsule on thoracolumbar burst fracture in senile patients with osteoporosis. Chinese Aubert, J., Begriche, K., Knockaert, L., Robin, M.A., Fromenty, B., 2011. Increased ex- J. Tradit. Med. Traumatol. Orthop. fi pression of cytochrome P450 2E1 in nonalcoholic fatty liver disease: mechanisms and Hernandez-Gea, V., Friedman, S.L., 2011. Pathogenesis of liver brosis. Annu. Rev. – pathophysiological role. Clin. Res. Hepatol. Gastroenterol. 35, 630–637. https://doi. Pathol. 6, 425 456. https://doi.org/10.1146/annurev-pathol-011110-130246. org/10.1016/j.clinre.2011.04.015. Honda, H., Ikejima, K., Hirose, M., Yoshikawa, M., Lang, T., Enomoto, N., Kitamura, T., fi Biesmans, S., Matthews, L.J., Bouwknecht, J.A., De Haes, P., Hellings, N., Meert, T.F., Takei, Y., Sato, N., 2002. Leptin is required for brogenic responses induced by – Nuydens, R., Ver Donck, L., 2016. Systematic analysis of the cytokine and anhedonia thioacetamide in the murine liver. Hepatology 36, 12 21. https://doi.org/https:// response to peripheral lipopolysaccharide administration in rats. BioMed Res. Int. doi.org/10.1053/jhep.2002.33684. 9085273. 2016. https://doi.org/10.1155/2016/9085273. Janssens, S., Beyaert, R., 2003. Role of Toll-like receptors in pathogen recognition. Clin. – Biesmans, S., Meert, T.F., Bouwknecht, J.A., Acton, P.D., Davoodi, N., De Haes, P., Microbiol. Rev. 16, 637 646. Kuijlaars, J., Langlois, X., Matthews, L.J., Ver Donck, L., Hellings, N., Nuydens, R., Johnston, D.E., 1999. Special considerations in interpreting liver function tests. Am. Fam. – 2013. Systemic immune activation leads to neuroinflammation and sickness behavior Physician 59, 2223 2230. in mice. Mediat. Inflamm. 271359. 2013. https://doi.org/10.1155/2013/271359. Kaneko, S., Satoh, T., Chiba, J., Ju, C., Inoue, K., Kagawa, J., 2000. Interleukin-6 and fl Brun, P., Castagliuolo, I., Di Leo, V., Buda, A., Pinzani, M., Palu, G., Martines, D., 2007. interleukin-8 levels in serum and synovial uid of patients with osteoarthritis. – Increased intestinal permeability in obese mice: new evidence in the pathogenesis of Cytokines Cell Mol. Ther. 6, 71 79. ff nonalcoholic steatohepatitis. Am. J. Physiol. Gastrointest. Liver Physiol. 292, Kanzler, H., Barrat, F.J., Hessel, E.M., Co man, R.L., 2007. Therapeutic targeting of in- G518–G525. https://doi.org/10.1152/ajpgi.00024.2006. nate immunity with Toll-like receptor agonists and antagonists. Nat. Med. 13, – Cao, Q., Mak, K.M., Lieber, C.S., 2007. Leptin represses matrix metalloproteinase-1 gene 552 559. https://doi.org/10.1038/nm1589. expression in LX2 human hepatic stellate cells. J. Hepatol. 46, 124–133. https://doi. Kim, H.A., Cho, M.L., Choi, H.Y., Yoon, C.S., Jhun, J.Y., Oh, H.J., Kim, H.Y., 2006. The org/10.1016/j.jhep.2006.07.027. catabolic pathway mediated by Toll-like receptors in human osteoarthritic chon- – Cederbaum, A.I., 2010. Role of CYP2E1 in ethanol-induced oxidant stress, fatty liver and drocytes. Arthritis Rheum. 54, 2152 2163. https://doi.org/10.1002/art.21951. hepatotoxicity. Dig. Dis. 28, 802–811. https://doi.org/10.1159/000324289. Kornaat, P.R., Bloem, J.L., Ceulemans, R.Y., Riyazi, N., Rosendaal, F.R., Nelissen, R.G., Cederbaum, A.I., Yang, L., Wang, X., Wu, D., 2012. CYP2E1 sensitizes the liver to LPS- Carter, W.O., Hellio Le Graverand, M.P., Kloppenburg, M., 2006. Osteoarthritis of the fi and TNF alpha-induced toxicity via elevated oxidative and nitrosative stress and knee: association between clinical features and MR imaging ndings. Radiology 239, – activation of ASK-1 and JNK mitogen-activated kinases. Int. J. Hepatol. 582790. 811 817. https://doi.org/10.1148/radiol.2393050253. 2012. https://doi.org/10.1155/2012/582790. Lewiecki, E.M., 2008. Prevention and treatment of postmenopausal osteoporosis. Obstet. – Chatterjee, S., Das, S., 2015. P2X7 receptor as a key player in oxidative stress-driven cell Gynecol. Clin. N. Am. 35, 301 315. ix. https://doi.org/10.1016/j.ogc.2008.03.007. fate in nonalcoholic steatohepatitis. Oxid. Med. Cell Longev. 172493. 2015. https:// Li, G., Hao, S., 2007. Application of Xianlinggubao in bone and arthrosis disease. Chin. J. doi.org/10.1155/2015/172493. N. Drugs Clin. Remedies. Chatterjee, S., Ganini, D., Tokar, E.J., Kumar, A., Das, S., Corbett, J., Kadiiska, M.B., Li, Z.R., Cheng, L.M., Wang, K.Z., Yang, N.P., Yang, S.H., He, W., Wang, Y.S., Wang, Z.M., Waalkes, M.P., Diehl, A.M., Mason, R.P., 2013. Leptin is key to peroxynitrite-medi- Yang, P., Liu, X.Z., Luo, Y.Z., Sun, W., Wang, H.T., Zheng, L.Z., Wang, X.L., Qin, L., ated oxidative stress and Kupffer cell activation in experimental non-alcoholic stea- 2018. Herbal Fufang Xian Ling Gu Bao prevents corticosteroid-induced osteonecrosis fi tohepatitis. J. Hepatol. 58, 778–784. https://doi.org/10.1016/j.jhep.2012.11.035. of the femoral head-A rst multicentre, randomised, double-blind, placebo-controlled – Chen, Y., Choi, S.S., Michelotti, G.A., Chan, I.S., Swiderska-Syn, M., Karaca, G.F., Xie, G., clinical trial. J. Orthop. Transl. 12, 36 44. https://doi.org/10.1016/j.jot.2017.11. 001.

10 W. Wu, et al. Journal of Ethnopharmacology 239 (2019) 111910

Lorenz, W., Buhrmann, C., Mobasheri, A., Lueders, C., Shakibaei, M., 2013. Bacterial C.E., Park, B.K., 2011. Circulating microRNAs as potential markers of human drug- lipopolysaccharides form procollagen-endotoxin complexes that trigger cartilage in- induced liver injury. Hepatology 54, 1767–1776. https://doi.org/10.1002/hep. flammation and degeneration: implications for the development of rheumatoid ar- 24538. thritis. Arthritis Res. Ther. 15, R111. https://doi.org/10.1186/ar4291. Tella, S.H., Gallagher, J.C., 2014. Prevention and treatment of postmenopausal osteo- Lou, H.Y., Huang, D., Fang, S.H., Zhao, X.F., Huang, Q.X., Wang, M.Z., 2009. Clinic re- porosis. J. Steroid Biochem. Mol. Biol. 142, 155–170. https://doi.org/10.1016/j. search on Xianling Gubao capsules for treatment of menopause syndrome. Zhongguo jsbmb.2013.09.008. Zhongyao Zazhi 34, 2950–2952. Thrailkill, K.M., Liu, L., Wahl, E.C., Bunn, R.C., Perrien, D.S., Cockrell, G.E., Skinner, R.A., Lu, J., Miyakawa, K., Roth, R.A., Ganey, P.E., 2013. Tumor necrosis factor-alpha po- Hogue, W.R., Carver, A.A., Fowlkes, J.L., Aronson, J., Lumpkin, C.K.J., 2005. Bone tentiates the cytotoxicity of amiodarone in Hepa1c1c7 cells: roles of caspase acti- formation is impaired in a model of type 1 diabetes. Diabetes 54, 2875–2881. vation and oxidative stress. Toxicol. Sci. 131, 164–178. https://doi.org/10.1016/j. Tian, F.M., Zhang, L., Luo, Y., Song, Y.Q., Jiao, A., 2011. Xianlinggubao promotes vas- taap.2012.11.00910.1093/toxsci/kfs289. cular formation in osteoporotic fracture callus. J. Clin. Rehabilitative Tissue Eng. Res. Maddox, J.F., Luyendyk, J.P., Cosma, G.N., Breau, A.P., Bible Jr., R.H., Harrigan, G.G., 15, 5161–5164. Goodacre, R., Ganey, P.E., Cantor, G.H., Cockerell, G.L., Roth, R.A., 2006. Tsuchida, T., Friedman, S.L., 2017. Mechanisms of hepatic stellate cell activation. Nat. Metabonomic evaluation of idiosyncrasy-like liver injury in rats cotreated with ra- Rev. Gastroenterol. Hepatol. 14, 397–411. https://doi.org/10.1038/nrgastro. nitidine and lipopolysaccharide. Toxicol. Appl. Pharmacol. 212, 35–44. https://doi. 2017.38. org/10.1016/j.taap.2005.06.021. Wang, K., Zhang, S., Marzolf, B., Troisch, P., Brightman, A., Hu, Z., Hood, L.E., Galas, Miller, M.A., McTernan, P.G., Harte, A.L., Silva, N.F., Strazzullo, P., Alberti, K.G., Kumar, D.J., 2009. Circulating microRNAs, potential biomarkers for drug-induced liver in- S., Cappuccio, F.P., 2009. Ethnic and sex differences in circulating endotoxin levels: a jury. Proc. Natl. Acad. Sci. U. S. A. 106, 4402–4407. https://doi.org/10.1073/pnas. novel marker of atherosclerotic and cardiovascular risk in a British multi-ethnic 0813371106. population. Atherosclerosis 203, 494–502. https://doi.org/10.1016/j.atherosclerosis. Wang, X., He, Y., Guo, B., Tsang, M.C., Tu, F., Dai, Y., Yao, Z., Zheng, L., Xie, X., Wang, N., 2008.06.018. 2015. In vivo screening for anti-osteoporotic fraction from extract of herbal formula Musumeci, G., Castrogiovanni, P., Trovato, F.M., Imbesi, R., Giunta, S., Szychlinska, M.A., xianlinggubao in ovariectomized mice. PLoS One 10, e0118184. Loreto, C., Castorina, S., Mobasheri, A., 2015. Physical activity ameliorates cartilage Wheeler, M.D., 2003. Endotoxin and Kupffer cell activation in alcoholic liver disease. degeneration in a rat model of aging: a study on lubricin expression. Scand. J. Med. Alcohol Res. Health 27, 300–306. Sci. Sports 25, e222–e230. https://doi.org/10.1111/sms.12290. Wu, D., Wang, X., Zhou, R., Cederbaum, A., 2010. CYP2E1 enhances ethanol-induced Musumeci, G., Trovato, F.M., Pichler, K., Weinberg, A.M., Loreto, C., Castrogiovanni, P., lipid accumulation but impairs autophagy in HepG2 E47 cells. Biochem. Biophys. 2013. Extra-virgin olive oil diet and mild physical activity prevent cartilage degen- Res. Commun. 402, 116–122. https://doi.org/10.1016/j.bbrc.2010.09.127. eration in an osteoarthritis model: an in vivo and in vitro study on lubricin expres- Wu, H., Zhong, Q., Wang, J., Wang, M., Fang, F., Xia, Z., Zhong, R., Huang, H., Ke, Z., sion. J. Nutr. Biochem. 24, 2064–2075. Wei, Y., Feng, L., Shi, Z., Sun, E., Song, J., Jia, X., 2017. Beneficial effects and toxicity Paik, Y.H., Kim, J., Aoyama, T., De Minicis, S., Bataller, R., Brenner, D.A., 2014. Role of studies of Xian-ling-gu-bao on bone metabolism in ovariectomized rats. Front. NADPH oxidases in liver fibrosis. Antioxidants Redox Signal. 20, 2854–2872. https:// Pharmacol. 8, 273. https://doi.org/10.3389/fphar.2017.00273. doi.org/10.1089/ars.2013.5619. Wu, J.J., Wen, L.P., Wu, Y.G., Shen, Q., Han, Y., 2009. [Effects of Xianling Gubao capsules Poulsen, K.L., Olivero-Verbel, J., 2014. Trovafloxacin Enhances Lipopolysaccharide- for the treatment of bone loss induced by glucocorticoid]. Zhong Guo Gu Shang 22, Stimulated Production of Tumor Necrosis Factor-Alpha by Macrophages: Role of the 193. DNA Damage Response, vol. 350. pp. 164–170. https://doi.org/10.1124/jpet.114. Yan, M., Zhang, J., Yang, H., Sun, Y., 2018. The role of leptin in osteoarthritis. Med 97, 214189. e0257. https://doi.org/10.1097/MD.0000000000010257. QI, X., REN, J., 2016. Comparison between water and ethanol extracts of xianlinggubao Yang, W.H., Liu, S.C., Tsai, C.H., Fong, Y.C., Wang, S.J., Chang, Y.S., Tang, C.H., 2013. for effects on hepatic mitochondria. Chinese J. Pharmacovigil. 13, 201–205. Leptin induces IL-6 expression through OBRl receptor signaling pathway in human Qiuqiong, L.I., 2010. Xianlinggubao capsule in promoting fracture healing efficacy. China synovial fibroblasts. PLoS One 8, e75551. https://doi.org/10.1371/journal.pone. Mod. Med. 0075551. Quach, D., Britton, R.A., 2017. Gut microbiota and bone health. Adv. Exp. Med. Biol. Yoshino, S., Sasatomi, E., Ohsawa, M., 2000. Bacterial lipopolysaccharide acts as an ad- 1033, 47–58. https://doi.org/10.1007/978-3-319-66653-2_4. juvant to induce autoimmune arthritis in mice. Immunology 99, 607–614. Raisz, L.G., 2005. Pathogenesis of osteoporosis: concepts, conflicts, and prospects. J. Clin. Zhang, G., Qin, L., Shi, Y., 2007. Epimedium-derived phytoestrogen flavonoids exert Investig. 115, 3318–3325. https://doi.org/10.1172/JCI27071. beneficial effect on preventing bone loss in late postmenopausal women: a 24-month Roth, R.A., Harkema, J.R., Pestka, J.P., Ganey, P.E., 1997. Is exposure to bacterial en- randomized, double-blind and placebo-controlled trial. J. Bone Miner. Res. 22, dotoxin a determinant of susceptibility to intoxication from xenobiotic agents? 1072–1079. https://doi.org/10.1359/jbmr.070405. Toxicol. Appl. Pharmacol. 147, 300–311. https://doi.org/10.1006/taap.1997.8301. Zhu, H.M., Qin, L., Garnero, P., Genant, H.K., Zhang, G., Dai, K., Yao, X., Gu, G., Hao, Y., Scanzello, C.R., Plaas, A., Crow, M.K., 2008. Innate immune system activation in os- Li, Z., Zhao, Y., Li, W., Yang, J., Zhao, X., Shi, D., Fuerst, T., Lu, Y., Li, H., Zhang, X., teoarthritis: is osteoarthritis a chronic wound? Curr. Opin. Rheumatol. 20, 565–572. Li, C., Zhao, J., Wu, Q., Zhao, S.J., 2012. The first multicenter and randomized https://doi.org/10.1097/BOR.0b013e32830aba34. clinical trial of herbal Fufang for treatment of postmenopausal osteoporosis. Schelbergen, R.F., Blom, A.B., van den Bosch, M.H., Sloetjes, A., Abdollahi-Roodsaz, S., Osteoporos. Int. 23, 1317–1327. https://doi.org/10.1007/s00198-011-1577-2. Schreurs, B.W., Mort, J.S., Vogl, T., Roth, J., van den Berg, W.B., van Lent, P.L., 2012. Zou, W., Devi, S.S., Sparkenbaugh, E., Younis, H.S., Roth, R.A., Ganey, P.E., 2009. Alarmins S100A8 and S100A9 elicit a catabolic effect in human osteoarthritic Hepatotoxic interaction of sulindac with lipopolysaccharide: role of the hemostatic chondrocytes that is dependent on Toll-like receptor 4. Arthritis Rheum. 64, system. Toxicol. Sci. 108, 184–193. https://doi.org/10.1093/toxsci/kfn259. 1477–1487. https://doi.org/10.1002/art.33495. Scotece, M., Conde, J., Lopez, V., Lago, F., Pino, J., Gomez-Reino, J.J., Gualillo, O., 2014. Adiponectin and leptin: new targets in inflammation. Basic Clin. Pharmacol. Toxicol. Abbreviations 114, 97–102. https://doi.org/10.1111/bcpt.12109. Sharapova, T., Devanarayan, V., LeRoy, B., Liguori, M.J., Blomme, E., Buck, W., Maher, XLGB: xian-ling-gu-bao J., 2016. Evaluation of miR-122 as a serum biomarker for hepatotoxicity in in- CYP2E1: cytochrome P450 2E1 – vestigative rat toxicology studies. Vet. Pathol. 53, 211 221. https://doi.org/10. NADPH: nicotinic adenine dinucleotide phosphate 1177/0300985815591076. iNOS: inducible nitric oxide synthase Shaw, P.J., Ditewig, A.C., Waring, J.F., Liguori, M.J., Blomme, E.A., Ganey, P.E., Roth, OA: Osteoarthritis fl R.A., 2009a. Coexposure of mice to trova oxacin and lipopolysaccharide, a model of BML: bone marrow lesions fi idiosyncratic hepatotoxicity, results in a unique gene expression pro le and inter- LPS: lipopolysaccharide – feron gamma-dependent liver injury. Toxicol. Sci. 107, 270 280. https://doi.org/10. OVX: ovariectomized 1093/toxsci/kfn205. SD: Sprague-Dawley Shaw, P.J., Ganey, P.E., Roth, R.A., 2009b. Tumor necrosis factor alpha is a proximal iDILI: idiosyncratic Drug-Induced Liver Injury fl mediator of synergistic hepatotoxicity from trova oxacin/lipopolysaccharide coex- HPLC: High performance liquid phase – posure. J. Pharmacol. Exp. Ther. 328, 62 68. https://doi.org/10.1124/jpet.108. SPF: specific pathogen free 143792. H&E: hematoxylin and eosin Shaw, P.J., Hopfensperger, M.J., Ganey, P.E., Roth, R.A., 2007. Lipopolysaccharide and ALT: alanine transaminase fl trova oxacin coexposure in mice causes idiosyncrasy-like liver injury dependent on AST: aspartate transaminase – tumor necrosis factor-alpha. Toxicol. Sci. 100, 259 266. https://doi.org/10.1093/ GGT: gamma-glutamyl transferase toxsci/kfm218. TBIL: total bilirubin Starckx, S., Batheja, A., Verheyen, G.R., Jonghe, S.D., Steemans, K., Dijck, B.V., Singer, TG: triglyceride M., Bogdan, N., Snoeys, J., Vinken, P., Sasaki, J.C., Gompel, J.V., Guzzie-Peck, P., miR-122: microRNA-122 Lampo, A., Lammens, L., 2013. Evaluation of miR-122 and other biomarkers in dis- IHC: Immunohistochemistry – tinct acute liver injury in rats. Toxicol. Pathol. 41, 795 804. https://doi.org/10. SEM: standard error of the mean 1177/0192623312464436. 3-NT: 3-Nitrotyrosine Starkey Lewis, P.J., Dear, J., Platt, V., Simpson, K.J., Craig, D.G., Antoine, D.J., French, N.S., Dhaun, N., Webb, D.J., Costello, E.M., Neoptolemos, J.P., Moggs, J., Goldring,

11