Recombinant human soluble thrombomodulin improved ... - Core

Journal of Pharmacological Sciences 129 (2015) 233e239

H O S T E D BY

Contents lists available at ScienceDirect

Journal of Pharmacological Sciences journal homepage: www.elsevier.com/locate/jphs

Full paper

Recombinant human soluble thrombomodulin improved lipopolysaccharide/D-galactosamine-induced acute liver failure in mice Wataru Osumi a, Denan Jin b, Yoshiro Imai a, Keitaro Tashiro a, Zhong-Lian Li c, Yoshinori Otsuki c, Kentaro Maemura c, Koji Komeda a, Fumitoshi Hirokawa a, Michihiro Hayashi a, Shinji Takai d, *, Kazuhisa Uchiyama a a

Department of General and Gastroenterological Surgery, Osaka Medical College, Takatsuki 569-8686, Japan Department of Pharmacology, Osaka Medical College, Takatsuki 569-8686, Japan c Department of Anatomy and Cell Biology, Osaka Medical College, Takatsuki 569-8686, Japan d Laboratory for Innovative Medicine, Graduate School of Medicine, Osaka Medical College, Takatsuki 569-8686, Japan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 July 2015 Received in revised form 30 August 2015 Accepted 18 November 2015 Available online 26 November 2015

The effect of recombinant human soluble thrombomodulin (TM-a) on acute liver failure (ALF) is unclear, and we elucidated the effect of TM-a in lipopolysaccharide (LPS)/D-galactosamine (GalN)-induced ALF in mice. Placebo (saline) or TM-a (100 mg/kg) was administered 1 h after LPS/GalN administration. Survival rates were evaluated for 24 h after LPS/GalN administration. Plasma and liver samples were evaluated 1, 3, and 7 h after LPS/GalN administration. Survival rates were significantly higher in the TM-a-treated group than in the placebo group. A significant augmentation of plasma high-mobility group box 1 protein (HMGB1) was observed 7 h after LPS/GalN administration. In the TM-a-treated mice, plasma HMGB1 was significantly lower than in the placebo group. A significant augmentation of hepatic nuclear factor (NF)kB p65 was observed in the placebo-treated group, whereas a significant reduction, relative to placebo, was observed in the TM-a-treated group. Hepatic expression of tumor necrosis factor (TNF)-a and myeloperoxidase were significantly increased in the placebo group, and were similarly significantly attenuated in the TM-a-treated group. TM-a treatment also produced a significant attenuation of liver neutrophil accumulation after LPS/GalN administration. Thus, TM-a may become a useful treatment strategy for reducing the symptoms of ALF via the attenuation of LPS/GalN-induced HMGB1 levels. © 2015 Japanese Pharmacological Society. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Acute liver failure High-mobility group box 1 protein Thrombomodulin Tumor necrosis factor-a Nuclear factor-kB

1. Introduction Acute liver failure (ALF) is a characterized by a rapid loss of hepatocyte function, and its major causes are viral infection, autoimmune hepatitis, and drug allergy (1). ALF is associated with high mortality in patients, and liver transplantation is the only effective therapeutic strategy for attenuating ALF. ALF is typically initiated by the activation of inflammatory cells such as macrophages, which release several inflammatory cytokines, resulting in massive death of parenchymal hepatocytes (2). * Corresponding author. Laboratory for Innovative Medicine, Graduate School of Medicine, Osaka Medical College, 2-7 Daigaku-machi, Takatsuki 569-8686, Japan. Tel.: þ81 72 684 6021; fax: þ81 72 684 6730. E-mail address: [email protected] (S. Takai). Peer review under responsibility of Japanese Pharmacological Society.

Combined administration of lipopolysaccharide (LPS) and Dgalactosamine (GalN) has become an established method for inducing ALF in experimental animals (3,4). LPS causes sinusoidal injury and intrahepatic fibrin deposition, which results in severe liver dysfunction (5). GalN potentiates the toxic effects of LPS in the liver through upregulation of toll-like receptor (TLR)-4 mRNA expression. TLR-4 specifically binds LPS in the liver (6). GalNpotentiated binding of LPS to TLR-4 in the liver promotes the activation of nuclear factor (NF)-kB, ultimately inducing the production of proinflammatory cytokines such as tumor necrosis factor (TNF)-a (7). Activation of NF-kB induces inflammation and apoptosis in the liver and increases TNF-a production, which exacerbates the inflammation (3). Consequently, the combined administration of LPS and GalN has been shown to specifically induce liver injury via activation of NF-kB and increased production of TNF-a (8).

http://dx.doi.org/10.1016/j.jphs.2015.11.007 1347-8613/© 2015 Japanese Pharmacological Society. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).

234

W. Osumi et al. / Journal of Pharmacological Sciences 129 (2015) 233e239

High-mobility group box 1 (HMGB1) is an intra-nuclear nonhistone DNA-binding protein that has both nuclear and extracellular functions (9). HMGB1 is secreted by necrotic and damaged cells, and release of HMGB1 into the extracellular space induces the production of proinflammatory cytokines such as TNF-a, resulting in a vicious cycle in which activated cytokines induce further release of HMGB1 (10,11). Upregulation of HMGB1 has been shown to be a lethal late-phase mediator of sepsis (12). Thrombomodulin (TM) is a thrombin receptor on the surface of endothelial cells and plays a crucial role in regulating coagulation (13). Additionally, TM activates protein C, resulting in antiinflammatory effects such as the suppression of TNF-a via inhibition of macrophage activation (14). In the present study, we used recombinant human soluble thrombomodulin (TM-a), which not only binds to HMGB1 but also cleaves HMGB1 by forming a complex with thrombin, resulting in attenuation of HMGB1-mediated inflammation (15). TM-a has proven to be useful in treating disseminated intravascular coagulation in clinic. In a rat LPSinduced sepsis model, TM-a was shown to significantly attenuate the upregulation of plasma HMGB1 levels (16). However, the role of TM-a in ALF remains unclear. In the present study, we evaluated the effect of TM-a on survival rate in a mouse model of LPS/GalNinduced ALF. 2. Materials and methods 2.1. Drugs LPS (Escherichia coli, 0111:B4) and GalN were purchased from Sigma (St. Louis, MO). TM-a was obtained from Asahi Kasei Pharma Co. Ltd. (Tokyo, Japan). 2.2. Animal model Eight-week-old male C57BL/6 mice (n ¼ 84) were obtained from Japan SLC (Shizuoka, Japan) and housed in a temperature-, humidity-, and light-controlled room. The effect of TM-a on the survival rate over the 24 h period following administration (i.p.) of LPS (4 mg/kg)/GalN (600 mg/kg) was evaluated by administering (s.c.) TM-a (100 mg/kg) or placebo (saline) 1 h after LPS/GalN injection (each group: n ¼ 20). Measurements of plasma and liver parameters were taken before and 1, 3, and 7 h after LPS/GalN injection (n ¼ 6) to assess time-dependent changes. The effects of TM-a were evaluated following administration (s.c.) of TM-a (100 mg/kg) or placebo (saline) 1 h after LPS/GalN injection, and blood and liver tissue were obtained at 7 h after LPS/GalN injection (n ¼ 8). Agematched normal mice were used as a control group (n ¼ 6). All procedures involving animals were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals at Osaka Medical College. 2.3. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activities in plasma Plasma was separated from the blood samples by centrifugation at 3000 rpm for 15 min at 4 C. Measurements of plasma activities of AST and ALT were performed by SRL Co. Ltd. (Tokyo, Japan).

2.5. Active NF-kB p65 in liver To measure NF-kB p65, nuclear proteins were isolated from fresh mouse liver using an NF-kB p65 ELISA kit (NOVUS Biologicals, Abingdon, UK) according to the manufacturer's instructions. Then, NF-kB p65 was measured using the NF-kB p65 ELISA kit. Protein concentrations of the nuclear fraction were assayed using the bicinchoninic acid Protein Assay Reagent (Pierce, Rockford, IL) with bovine serum albumin as the standard. 2.6. Real-time polymerase chain reaction (RT-PCR) Liver total RNA was extracted using the Trizol reagent (Life Technologies, Rockville, MD) and subsequently dissolved in RNasefree water (Takara Bio Inc. Otsu, Japan). Total RNA (1 mg) was transcribed into cDNA with Superscript VIRO (Invitrogen, Carlsbad, CA). Levels of mRNA were measured by RT-PCR on a LightCycler with software (Roche Diagnostics, Tokyo, Japan) using TaqMan fluorogenic probes. Primers and probes for RT-PCR of TNF-a, myeloperoxidase (MPO), and 18S ribosomal RNA (rRNA) were designed by Roche Diagnostics. The primers were as follows: 50 -aggcgaagattactgccaag-30 (forward) and 50 -catggctatgaggtagagacagg-30 (reverse) for TNF-a, 50 -ctgaatcctcgatggaatgg-30 (forward) and 50 ccatggcccctacaatctt-30 (reverse) for MPO, and 50 -gcaattattccccatgaacg-30 (forward) and 50 -gggacttaatcaacgcaagc-30 (reverse) for 18S rRNA. The probes were as follows: 50 -agccccag-30 for TNF-a, 50 ccaggagg-30 for MPO, and 50 -ttcccagt-30 for 18S rRNA. mRNA levels of TNF-a and MPO were normalized to that of 18S rRNA. 2.7. Histological analysis The liver tissue specimens were fixed with Carnoy's fixative in 10% methanol overnight. The fixed liver tissues were embedded in paraffin, and then cut at a thickness of 5 mm. The sections were mounted on silanized slides (Matsunami, Kishiwada, Japan) and deparaffinized with xylene and ethanol. The severity of hepatic histological changes was assessed using hematoxylin and eosin (HE) staining. The procedure for immunohistochemical analysis of MPO has been previously described (17). Sections were incubated with antiMPO antibody (Abcam Inc. Cambridge, MA) followed by a reaction with appropriate reagents from a streptavidin-biotin peroxidase kit (Dako LSABkit; Dako Co., Carpinteria, CA) and 3-amino-9ethylcarbazole, which was used for color development. The sections were lightly counterstained with hematoxylin. 2.8. Statistical analysis Data are expressed as the mean ± standard error of the mean (SEM). Significant differences among mean values of multiple groups were evaluated using a one-way analysis of variance followed by Fisher's test. The cumulative survival in each group was determined using the KaplaneMeier method, and survival was compared between groups with a log-rank test. Values of P < 0.05 were considered statistically significant. 3. Results 3.1. Effect of TM-a on survival rate after LPS/GalN administration

2.4. HMGB1 in plasma HMGB1 in plasma was measured with an enzyme-linked immunosorbent assay (ELISA) kit (Shino-test Co., Sagamihara, Japan) according to the manufacturer's instructions.

Twenty-four hours after LPS/GalN administration, the survival rates were 15% and 55% in the placebo- and TM-a-treated groups, respectively (Fig. 1). However, additional mortality was not observed in either group after 24 h (Fig. 1). The survival rates


3.3. Effect of TM-a on HMGB1 in plasma after LPS/GalN administration

100 80

Plasma HMGB1 tended to be higher 3 h after LPS/GalN administration, but did not reach significance (Fig. 3A). However, a significant increase in plasma HMGB1 was observed at 7 h (Fig. 3A). The significant increase in HMGB1 observed in the placebo-treated group 7 h after LPS/GalN administration was significantly attenuated by TM-a treatment (Fig. 3B).

60 40

3.4. Effect of TM-a on NF-kB p65 in liver after LPS/GalN administration

0 0

2

4

6

8 10 12 14 16 18 20 22 24

Time after LPS/GalN(h) Fig. 1. Cumulative percent survival following LPS/GalN administration in placebo (solid line) and TM-a (broken line) treated animals. The difference in percent survival between placebo- and TM-a-treated animals was statistically significant (KaplaneMeier analysis followed by log rank test; P < 0.01; n ¼ 20).

between the placebo- and TM-a-treated groups were significantly different (Fig. 1). 3.2. Effect of TM-a on AST and ALT activities in plasma after LPS/ GalN administration Time courses of AST and ALT activities in plasma after LPS/GalN administration are shown in Fig. 2A and C. Both AST and ALT activities in plasma were not significantly altered until 3 h after LPS/ GalN administration, compared with the pre-administration plasma levels; however, activities were significantly increased at 7 h (Fig. 2A and C). To evaluate the effect of TM-a on the activity of AST and ALT in plasma, we administered TM-a 1 h after LPS/GalN treatment and evaluated activity at 7 h, when significant increases in post-LPS/ GalN activities were observed. Both AST and ALT activities were significantly lower in the TM-a-treated group compared to the placebo-treated group (Fig. 2B and D).

ALT (U/L)

AST (U/L)

A 2000 1600 1200 800 400 0

**

A significant increase in liver NF-kB p65 was observed 1, 3, and 7 h after LPS/GalN administration compared with preadministration level (Fig. 4A). TM-a treatment significantly attenuated NF-kB p65 in liver compared with the placebo-treated group 7 h after LPS/GalN administration (Fig. 4B). 3.5. Effects of TM-a on hepatic TNF-a and MPO gene expression after LPS/GalN administration Hepatic gene expression of TNF-a was significantly higher 1 h after LPS/GalN administration compared with the preadministration level (Fig. 5A). Augmented TNF-a gene expression was significantly attenuated in the TM-a-treated group (Fig. 5B). Hepatic MPO gene expression was not significantly different up to 3 h after LPS/GalN administration compared with the preadministration level, while a significant augmentation of MPO gene expression was observed at 7 h (Fig. 5C). MPO gene expression was significantly attenuated by TM-a treatment (Fig. 5D). 3.6. Effects of TM-a on MPO-positive cell numbers after LPS/GalN administration Representative images of HE-stained liver sections from control, placebo-, and TM-a-treated mice 7 h after LPS/GalN administration are shown in Fig. 6A. Numerous necrotic lesions were observed in the placebo-treated group; however, these lesions were clearly reduced in the TM-a-treated

B AST (U/L)

20

Pre 1 3 7 Time after LPS/GalN (h)

2000 1600 1200 800 400 0

C

D

1000 800 600 400 200 0

1000 800 600 400 200 0

**


ALT (U/L)

Survival (%)

235

**

Control

**

TM-α

Placebo

LPS/GalN

**

**

Control

Placebo

TM-α LPS/GalN

Fig. 2. Time course of plasma AST (A) and ALT (C) activities before (Pre) and 1, 3, and 7 h after LPS/GalN administration. Values represent mean ± SEM (n ¼ 6) (A and C). **P < 0.01 vs. pre-administration (A and C). Plasma AST (B) and ALT (D) activities in control and placebo- or TM-a-treated animals 7 h after LPS/GalN administration. Values represent mean ± SEM (n ¼ 8) (B and D). **P < 0.01 vs. placebo-treated group (B and D).

236


B

**

30 20 10 0

HMGB1 (ng/mL)

HMGB1 (ng/mL)

A 40

**

30 20 10 0


**

40

Control

Placebo

TM-α

LPS/GalN

Fig. 3. Time course of plasma HMGB1 levels before (Pre) and 1, 3, and 7 h after LPS/GalN administration (A). Values represent mean ± SEM (n ¼ 6) (A). **P < 0.01 vs. preadministration (A). Plasma HMGB1 levels in control and placebo- or TM-a-treated animals 7 h after LPS/GalN administration (B). Values represent mean ± SEM (n ¼ 8) (B). **P < 0.01 vs. placebo-treated group (B).

B

1.5 1.0

*

*

*

0.5 0

NF-κB p65 (ng/mg protein)

NF-κB p65 (ng/mg protein)

A


*

**

0.8 0.6 0.4 0.2 0

Control

Placebo

TM-α

LPS/GalN

Fig. 4. Time course of hepatic NF-kB p65 before (Pre) and 1, 3, and 7 h after LPS/GalN administration (A). Values represent mean ± SEM (n ¼ 6) (A). *P < 0.05 vs. pre-administration (A). Hepatic NF-kB p65 levels in control and placebo- or TM-a-treated animals 7 h after LPS/GalN administration (B). Values represent mean ± SEM (n ¼ 8) (B). *P < 0.05 and **P < 0.01 vs. placebo-treated group (B).

*

1.0 0.5 0

C MPO/18S rRNA

*

1.5

1.0 0.8 0.6 0.4 0.2 0


**


TNF-α/18S rRNA

2.0

B 2.0

*

1.5

*

1.0 0.5 0

Control

Placebo

TM-α

LPS/GalN

D

MPO/18S rRNA

TNF-α/18S rRNA

A

1.0 0.8 0.6 0.4 0.2 0

*

Control

*

Placebo

TM-α

LPS/GalN

Fig. 5. Time course of hepatic gene expression of TNF-a (A) and MPO (C) before (Pre) and 1, 3, and 7 h after LPS/GalN administration. Values represent mean ± SEM (n ¼ 6) (A and C). *P < 0.05 and **P < 0.01 vs. pre-administration (A and C). Hepatic gene expression of TNF-a (B) and MPO (D) in control and placebo- or TM-a-treated animals 7 h after LPS/GalN administration. Values represent mean ± SEM (n ¼ 8) (B and D). *P < 0.05 vs. placebo-treated group (B and D).

group (Fig. 6A). Representative images of liver sections immunostained with anti-MPO antibody from control, placebo-, and TM-a-treated mice 7 h after LPS/GalN administration are shown in Fig. 6A.

The number of MPO-positive cells was significantly higher in the placebo-treated group than in the control group 7 h after LPS/GalN administration, but was significantly attenuated in the TM-atreated group (Fig. 6B).


B

HE (×100)

25

MPO-positive cells (cells/mm2)

A

MPO (×100)

MPO (×400)

Control

Placebo LPS/GalN

TM-α

**

237

**

20 15 10 5 0

Control Placebo TM-α LPS/GalN

Fig. 6. Representative images of HE-stain (upper images) and MPO-positive cells (middle and bottom images) in the liver sections from control, placebo-, and TM-a-treated mice 7 h after LPS/GalN administration (A). The original magnification was 100; scale bars are 100 mm (A: upper and middle images). The original magnification was 400; scale bars are 50 mm (A: lower images). MPO-positive cells in control, placebo-, and TM-a-treated animals 7 h after LPS/GalN administration (B). Values represent mean ± SEM (n ¼ 8). **P < 0.01 vs. placebo-treated group.

4. Discussion In the present study, we demonstrated for the first time that delayed (1 h post-LPS/GalN) administration of TM-a significantly reduced liver injury as well as survival rate in LPS/GalN-induced ALF. The importance of TNF-a has been demonstrate in the pathogenesis of ALF after LPS/GalN administration (4,18), and we observed a significant augmentation of hepatic TNF-a gene expression 1 h after LPS/GalN administration. LPS is potentiated by GalN in liver and immediately triggers the TLR-4 signaling pathway in macrophages, which activates NF-kB and results in the production of proinflammatory cytokines such as TNF-a (7). TNF-a promotes hepatic apoptosis via inflammation and oxidative stress (19). Numerous papers have demonstrated that pretreatment with antiinflammatory agents and antioxidants ameliorates the damage associated with liver injury and enhances the survival rate after LPS/GalN administration via a reduction in TNF-a levels (4,18,20). These findings illustrate the importance of the rapid elevation in TNF-a levels in the pathogenesis of ALF after LPS/GalN administration. However, the agents used in these previous studies were administered before LPS/GalN administration to attenuate the rapid elevation in TNF-a levels occurring after LPS/GalN administration. Prior studies have not determined whether attenuation of the initial increase in TNF-a levels after LPS/GalN administration is necessary to prevent ALF. In the present study, we found that the increases in both the activity of AST and ALT and mortality rate after LPS/GalN administration were significantly attenuated in TM-atreated mice, even though TM-a was administered 1 h after LPS/ GalN administration. These findings clearly demonstrate that treatment with TM-a after LPS/GalN administration can ameliorate liver injury and enhance survival in mice experiencing ALF, even in the presence of elevated TNF-a levels. A significant elevation of hepatic TNF-a gene expression has previously been observed 1 h after LPS/GalN administration (4,18). Although the significant elevation of TNF-a was not observed 3 h after LPS/GalN administration, significantly elevated levels were once again observed at 7 h. The mechanism underlying the biphasic elevation of TNF-a levels after LPS/GalN administration remains unclear. The initial increase in TNF-a level 1 h after LPS/GalN administration is thought to be dependent on direct stimulation of TLR-4 by LPS in liver macrophages, which in turn promotes the

secretion of proinflammatory cytokines such as TNF-a, leading to significant necrosis of hepatocytes (21). HMGB1 is released by necrotic and damaged cells (10,11), and the observed increase in HMGB1 levels at 7 h in the present study may be indicative of hepatic necrosis. In general, increased activity of AST and ALT reflects liver damage, and the activity of these enzymes was dramatically elevated 7 h after LPS/GalN administration. The second peak in TNF-a levels, observed 7 h after LPS/GalN administration, may be closely related to elevated HMGB1 levels in damaged liver. HMGB1 acts as an inflammatory signal that promotes the production of inflammatory cytokines such as TNF-a via the activation of p38 mitogen-activated protein kinase (22). In contrast, TNF-a may also induce the release of HMGB1 from damaged liver cells (12). Treatment with anti-TNF-a antibodies has been shown to ameliorate liver injury and increase the survival rate in mice in a manner that corresponds with the reduction in HMGB1 levels after LPS/GalN administration (23). On the other hand, in rats, treatment with anti-HMGB1 antibodies also attenuates TNF-a expression after LPS/GalN administration (24). In animal models, TM-a has been shown to diminish HMGB1 activity (15,25). Therefore, an increase in HMGB1 release from damaged liver cells may induce the production of TNF-a, which in turn induces HMGB1 release 7 h after LPS/GalN administration. Therefore, TM-a may interrupt the destructive cycle of TNF-a and HMGB1, preventing the progression of ALF. In addition to the increased inflammation and oxidative stress during the acute phase of LPS/GalN-induced damage, TNF-a contributes to neutrophil accumulation in the liver, which causes massive hepatocyte necrosis during the late phase pathology after LPS/GalN administration (25). Furthermore, HMGB1 also induces the accumulation of neutrophils through the upregulation of adhesion molecules, such as intercellular adhesion molecule-1 (26). The accumulation of neutrophils may be involved in the acceleration of late-phase ALF after LPS/GalN administration. In the present study, a significant increase in MPO activity was observed 7 h after LPS/GalN administration. Although monocytes/macrophages are fewer in number compared with neutrophils, monocytes/macrophages also contain MPO-positive granules (27). In the present study, the number of MPO-positive cells increased after LPS/GalN injection, and this increase was thought to be composed in part of monocytes/macrophages. MPO is known to induce inflammation

238


and oxidative stress, leading to hepatic necrosis (28). In a rat monocrotaline-induced sinusoidal obstruction syndrome model, TM-a was shown to increase the survival rate after monocrotaline administration due to attenuation of hepatic high-mobility group box 1 (HMGB1) protein levels, MPO activity, and neutrophil accumulation, all of which were significantly augmented 6 h after monocrotaline administration (29). HMGB1 was shown to promote the recruitment of neutrophils (but not macrophages) into necrotic liver tissues in mice, and attenuation of neutrophils (but not macrophages) was observed in mice in which the gene encoding HMGB1 was deleted (30). Therefore, the attenuation of MPOpositive cells by TM-a may primarily involve a decrease in the accumulation of neutrophils. However, to the best of our knowledge, it remains unclear whether neutrophil-neutralizing antibodies or neutrophil extracellular traps directly protect against liver injury after LPS/GalN injection. Further studies will be needed to investigate the relationship between neutrophil attenuation and liver protection in ALF. Activation of NF-kB is triggered by many factors, including LPS and TNF-a. The binding of LPS by TLR-4 induces translocation from the cytoplasm to the nucleus, where TLR-4 binds to NF-kB promotor sites and activates transcription of TNF-a (31). The binding TNF-a to its main receptor TNF-R1 also induces nuclear translocation and transcriptional activation of NF-kB-dependent target genes (8). NFkB activity is controlled by a family of cytoplasmic inhibitory molecules (I-kBs), and the p65:p50 NF-kB heterodimer is formed after dissociation from I-kBa, which is degraded by the 26S proteasome (8,32). The p65:p50 heterodimer is a major NF-kB DNA binding complex induced by LPS and TNF-a, and p65 has been used a major protein to monitor NF-kB activation. In the present study, we also measured NF-kB p65 in liver tissues, and observed a significant increase in NF-kB even 1 h after LPS/GalN administration. While the increase in NF-kB p65 continued to be observed for up to 7 h, the levels were significantly attenuated by post-treatment with TM-a. In rats, post-treatment with anti-HMGB1 antibodies after LPS/GalN administration significantly attenuated LPS/GalNinduced late-phase levels of both TNF-a and NF-kB p65 (24). In the present study, post-treatment with TM-a attenuated not only HMGB1 levels, but also TNF-a and NF-kB p65 levels. Thus, the mechanism underlying the beneficial effect produced by posttreatment with TM-a after LPS/GalN administration may be closely related to the attenuation of LPS/GalN-induced late-phase elevation of HMGB1. In the present study, we evaluated each parameter in mice administered TM-a at a dose of 100 mg/kg (s.c.). As indicated above, survival rates of 20 and 30% 24 h after LPS/GalN administration were observed in mice administered TM-a (s.c.) at doses of 10 and 30 mg/kg, respectively, and no significant difference was observed compared with placebo-treated mice (survival rate of 15%). At a 100 mg/kg dose of TM-a, the survival rate was 55%, and a significant difference was observed compared with placebo-treated mice, as shown in Fig. 1. A significant increase in survival rate (60%) compared with placebo was also observed at a TM-a (s.c.) dose of 200 mg/kg, but no significant difference was observed between mice receiving a TM-a (s.c.) dose of 100 or 200 mg/kg (data not shown). Therefore, we selected a TM-a dose of 100 mg/kg for evaluation of the various parameters. On the other hand, previous papers have reported anti-inflammatory effects of TM-a at lower doses. For example, 1 mg/kg of TM-a (i.v.) showed a significant effect in preventing lung injury in a rat model of LPS-induced systemic inflammation (16), and 3 mg/kg of TM-a (i.v.) showed a significant effect in preventing liver damage in a rat model of monocrotaline-induced sinusoidal obstruction syndrome (29). In a mouse heat stroke model, significant amelioration of liver injury was observed with administration of 1 mg/kg of TM-a (i.p) (33).

Although it is unclear why higher doses were required in the present study compared with previous studies, the route of administration (e.g., subcutaneous) may have affected our results. It can be difficult to achieve high blood concentrations of high-molecularweight substances such as TM-a (molecular weight approximately 64,000 Da) via subcutaneous injection, and this may have played a role in our study. Another possibility is that the pathogenesis of the animal model we used differs from that of other models. In conclusion, TM-a post-treatment attenuated liver damage and improved survival rates following LPS/GalN-induced ALF in mice, even when the levels of the proinflammatory cytokine TNF-a were elevated. The mechanism underlying these effects of TM-a may be mediated by the attenuation of LPS/GalN-induced HMGB1 levels during the late phase. Conflict of interest The authors declare that there are no conflicts of interest. Acknowledgment This study was partially supported by Grants-in-Aid for Scientific Research (C) of 26462049 (KK), 26461932 (FH), 26462050 (MH) and 15K08251 (ST) by from the Japan Society for the Promotion of Science. References (1) Bernal W, Auzinger G, Dhawan A, Wendon J. Acute liver failure. Lancet. 2010;376:190e201. (2) Wolf AM, Wolf D, Rumpold H, Ludwiczek S, Enrich B, Gastl G, et al. The kinase inhibitor imatinib mesylate inhibits TNF-a production in vitro and prevents TNF- dependent acute hepatic inflammation. Proc Natl Acad Sci USA. 2005;102:13622e13627. (3) Galanos C, Freudenberg MA, Reutter W. Galactosamine-induced sensitization to the lethal effects of endotoxin. Proc Natl Acad Sci USA. 1979;76: 5939e5943. (4) Imai Y, Takai S, Jin D, Komeda K, Tashiro K, Li ZL, et al. Chymase inhibition attenuates lipopolysaccharide/d-galactosamine-induced acute liver failure in hamsters. Pharmacology. 2014;93:47e56. (5) Nakao A, Taki S, Yasui M, Kimura Y, Nonami T, Harada A, et al. The fate of intravenously injected endotoxin in normal rats and in rats with liver failure. Hepatology. 1994;19:1251e1256. (6) Kitazawa T, Tsujimoto T, Kawaratani H, Fujimoto M, Fukui H. Expression of Toll-like receptor 4 in various organs in rats with D-galactosamine-induced acute hepatic failure. J Gastroenterol Hepatol. 2008;23:e494ee498. (7) Schwabe RF, Seki E, Brenner DA. Toll-like receptor signaling in the liver. Gastroenterology. 2006;130:1886e1900. (8) Luedde T, Trautwein C. Intracellular survival pathways in the liver. Liver Int. 2006;26:1163e1174. (9) Bustin M. Regulation of DNA-dependent activities by the functional motifs of the high-mobility-group chromosomal proteins. Mol Cell Biol. 1999;19: 5237e5246. (10) Wang H, Bloom O, Zhang M, Vishnubhakat JM, Ombrellino M, Che J, et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science. 1999;285: 248e251. (11) Scaffidi P, Misteli T, Bianchi ME. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature. 2002;418:191e195. (12) Wang H, Yang H, Czura CJ, Sama AE, Tracey KJ. HMGB1 as a late mediator of lethal systemic inflammation. Am J Respir Crit Care Med. 2001;164: 1768e1773. (13) Esmon CT. The interactions between inflammation and coagulation. Br J Haematol. 2005;131:417e430. (14) Grey ST, Tsuchida A, Hau H, Orthner CL, Salem HH, Hancock WW. Selective inhibitory effects of the anticoagulant activated protein C on the responses of human mononuclear phagocytes to LPS, IFN-gamma, or phorbol ester. J Immunol. 1994;153:3664e3672. (15) Ito T, Kawahara K, Okamoto K, Yamada S, Yasuda M, Imaizumi H, et al. Proteolytic cleavage of high mobility group box 1 protein by thrombinthrombomodulin complexes. Arterioscler Thromb Vasc Biol. 2008;28: 1825e1830. (16) Hagiwara S, Iwasaka H, Matsumoto S, Hasegawa A, Yasuda N, Noguchi T. In vivo and in vitro effects of the anticoagulant, thrombomodulin, on the inflammatory response in rodent models. Shock. 2010;33:282e288.

W. Osumi et al. / Journal of Pharmacological Sciences 129 (2015) 233e239 (17) Masubuchi S, Takai S, Jin D, Tashiro K, Komeda K, Li ZL, et al. Chymase inhibitor ameliorates hepatic steatosis and fibrosis on established non-alcoholic steatohepatitis in hamsters fed a methionine- and choline-deficient diet. Hepatol Res. 2013;43:970e978. (18) Kim SJ, Kim JK, Lee DU, Kwak JH, Lee SM. Genipin protects lipopolysaccharideinduced apoptotic liver damage in D-galactosamine-sensitized mice. Eur J Pharmacol. 2010;635:188e193. (19) Han D, Hanawa N, Saberi B, Kaplowitz N. Hydrogen peroxide and redox modulation sensitize primary mouse hepatocytes to TNF-induced apoptosis. Free Radic Biol Med. 2006;41:627e639. (20) Kim SH, Kim YS, Kang SS, Bae K, Hung TM, Lee SM. Anti-apoptotic and hepatoprotective effects of gomisin A on fulminant hepatic failure induced by Dgalactosamine and lipopolysaccharide in mice. J Pharmacol Sci. 2008;106: 225e233. (21) Zhan Y, Wang Z, Yang P, Wang T, Xia L, Zhou M, et al. Adenosine 50 -monophosphate ameliorates D-galactosamine/lipopolysaccharide-induced liver injury through an adenosine receptor-independent mechanism in mice. Cell Death Dis. 2014;5:e985. (22) Huang H, Evankovich J, Yan W, Nace G, Zhang L, Ross M, et al. Endogenous histones function as alarmins in sterile inflammatory liver injury through Toll-like receptor 9 in mice. Hepatology. 2011;54:999e1008. (23) Yang F, Li X, Wang LK, Wang LW, Han XQ, Zhang H, et al. Inhibitions of NF-kB and TNF-a result in differential effects in rats with acute on chronic liver failure induced by d-Gal and LPS. Inflammation. 2014;37: 848e857. (24) Li X, Wang LK, Wang LW, Han XQ, Yang F, Gong ZJ. Blockade of high-mobility group box-1 ameliorates acute on chronic liver failure in rats. Inflamm Res. 2013;62:703e709.

239

(25) Kawaguchi K, Kikuchi S, Hasegawa H, Maruyama H, Morita H, Kumazawa Y. Suppression of lipopolysaccharide-induced tumor necrosis factor-release and liver injury in mice by naringin. Eur J Pharmacol. 1999;368:245e250. (26) Feng L, Zhu M, Zhang M, Jia X, Cheng X, Ding S, et al. Amelioration of compound 4,40 -diphenylmethane-bis(methyl)carbamate on high mobility group box1-mediated inflammation and oxidant stress responses in human umbilical vein endothelial cells via RAGE/ERK1/2/NF-kB pathway. Int Immunopharmacol. 2013;15:206e216. (27) Klebanoff SJ. Myeloperoxidase: friend and foe. J Leukoc Biol. 2005;77: 598e625. (28) Malle E, Furtmüller PG, Sattler W, Obinger C. Myeloperoxidase: a target for new drug development? Br J Pharmacol. 2007;152:838e854. (29) Nakamura K, Hatano E, Miyagawa-Hayashino A, Okuno M, Koyama Y, Narita M, et al. Soluble thrombomodulin attenuates sinusoidal obstruction syndrome in rat through suppression of high mobility group box 1. Liver Int. 2014;34:1473e1487. (30) Huebener P, Pradere JP, Hernandez C, Gwak GY, Caviglia JM, Mu X, et al. The HMGB1/RAGE axis triggers neutrophil-mediated injury amplification following necrosis. J Clin Invest. 2015;125:539e550. € O, Brittenham G, Tsukamoto H. Role of iron in NF(31) Lin M, Rippe RA, Niemela kappa B activation and cytokine gene expression by rat hepatic macrophages. Am J Physiol. 1997;272:G1355eG1364. (32) Paik YH, Schwabe RF, Bataller R, Russo MP, Jobin C, Brenner DA. Toll-like receptor 4 mediates inflammatory signaling by bacterial lipopolysaccharide in human hepatic stellate cells. Hepatology. 2003;37:1043e1055. (33) Kawasaki T, Okamoto K, Kawasaki C, Sata T. Thrombomodulin improved liver injury, coagulopathy, and mortality in an experimental heatstroke model in mice. Anesth Analg. 2014;118:956e963.