Rho-kinase–dependent pathway mediates the hepatoprotective effects of sorafenib against ischemia/reperfusion liver injury in rats with nonalcoholic steatohepatitis

Authors

  • Ying-Ying Yang,

    Corresponding author
    1. Division of General Medicine, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan
    2. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    • Division of General Medicine, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan 11217
    Search for more papers by this author
    • Telephone: +886 02 28757506; FAX: +886-2-28739318

  • Yi-Tsau Huang,

    1. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    2. National Research Institute of Chinese Medicine, Taipei, Taiwan
    Search for more papers by this author
  • Tzung-Yan Lee,

    1. Graduate Institute of Traditional Chinese Medicine, Chang Gung University, Taipei, Taiwan
    Search for more papers by this author
  • Che-Chang Chan,

    1. Divisions of Gastroenterology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan
    2. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    Search for more papers by this author
  • Yi-Chen Yeh,

    1. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    2. Department of Pathology, Taipei Veterans General Hospital, Taipei, Taiwan
    Search for more papers by this author
  • Kuei-Chuan Lee,

    1. Divisions of Gastroenterology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan
    2. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    Search for more papers by this author
  • Han-Chieh Lin

    Corresponding author
    1. Divisions of Gastroenterology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan
    2. Department of Medicine, National Yang-Ming University School of Medicine, Taipei, Taiwan
    • Division of General Medicine, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan 11217
    Search for more papers by this author
    • Telephone: +886 02 28757506; FAX: +886-2-28739318


  • This study was supported by grants from the National Science Council (NSC100-2314-B-075-058-MY3 and NSC100-2314-B-075-015) and Taipei Veterans General Hospital (V100C-035 and VGH100C-21).

Abstract

During liver transplantation, nonalcoholic steatohepatitis (NASH) aggravates ischemia/reperfusion (IR) injury by activating various kinases and subsequently releasing cytokines and chemokines. Nonetheless, the effect of the multikinase inhibitor sorafenib on IR liver injury in rats with NASH has never been explored. Our study was designed to determine this effect and associated mechanisms in NASH rats. Sorafenib was acutely administered to NASH rats with IR liver injury that were or were not chronically pretreated with the Rho-kinase–specific inhibitor fasudil. Then, the following were evaluated: mean arterial pressure; hepatic blood flow and microcirculatory dysfunction; hepatic inflammation (serum alanine aminotransferase); necrosis; apoptosis; leukocyte infiltration; CD45 staining; caspase levels and DNA fragmentation; various serum and hepatic cytokines; and proteins and genes of the Raf/mitogen-activated protein–extracellular signal-regulated kinase kinase (MEK)/extracellular signal-regulated kinase (ERK), p38 mitogen-activated protein kinase, and apoptosis pathways. In NASH rats with IR liver injury, hepatic inflammation, necrosis, apoptosis, leukocyte infiltration, and microcirculatory dysfunction were significantly attenuated by the acute administration of sorafenib through the inhibition of the hepatic release of macrophage inflammatory protein 2, keratinocyte chemoattractant, granulocyte-monocyte colony-stimulating factor, and hepatic caspase-3 and caspase-9 as well as DNA fragmentation. Furthermore, there was decreased expression of p-Raf1 (where p indicates the phosphorylated form), p-MEK1/2, p-ERK1/2, p-Rho-kinase, B cell lymphoma 2–associated death promoter, and B cell lymphoma 2–associated X protein at the protein and messenger RNA levels. Notably, the aforementioned beneficial effects of sorafenib were significantly abolished by chronic pretreatment with the Rho-kinase–specific inhibitor fasudil. This study demonstrated that the multikinase inhibitor sorafenib protects NASH rats from IR injury by interfering with the inflammation, necrotic, and apoptotic responses causing leukocyte-dependent hepatic microcirculatory dysfunction. The hepatoprotective effects of sorafenib seem to work through the inhibition of the Rho-kinase–dependent Raf/MEK/ERK pathway, which is up-regulated during IR injury in the livers of NASH rats. Liver Transpl 18:1371–1383, 2012. © 2012 AASLD.

Ischemia/reperfusion (IR) liver injury is a major determinant of intraoperative and postoperative allograft dysfunction and morbidity.1 Livers with nonalcoholic steatohepatitis (NASH) are especially sensitive to IR injury.2 The need to protect donor liver grafts with NASH against IR injury is highlighted by the shortage of organs that would result if the pool of steatotic livers were avoided for transplantation.3 Currently, no effective approaches are available for managing NASH patients with IR-induced liver damage.

Generally, IR liver injury is linked to leukocyte activation and recruitment, extensive hepatic inflammation, necrosis, and apoptosis.2, 4-6 In these livers, leukocyte activation and recruitment begin with the release of cytokines and chemokines, including monocyte chemoattractant protein 2 (MCP1), macrophage inflammatory protein 2 (MIP2), and keratinocyte chemoattractant (KC).6 Thus, it is important to explore therapeutic agents that are able to limit this initial release of inflammatory cytokines and chemokines and thus prevent progression to irreversible liver injury.2, 5, 6

Sorafenib down-regulates the mitogen-activated protein kinase (MAPK) intracellular/extracellular signal transduction pathway, which comprises Raf1, mitogen-activated protein–extracellular signal-regulated kinase kinase (MEK), extracellular signal-regulated kinase (ERK), p38 mitogen-activated protein kinase (p38MAPK), and Rho-kinase.7, 8 The Raf/MEK/ERK pathway regulates the production of hepatic chemokines, leukocyte recruitment, and apoptosis during endotoxemia liver injury.9 Rho-kinase stimulates the release of interleukin-8 (IL-8) and MCP1 via the p38MAPK pathway.10 Conversely, the expression of Rho-kinase is up-regulated by inflammatory stimuli and apoptosis.11 Actually, Rho-kinase is involved in both apoptosis and necrosis in livers with IR injury.12, 13 Increased susceptibility to IR injury is a Rho/Rho-kinase–dependent process in steatotic livers.13

Accordingly, sorafenib is potentially a drug that might protect livers from IR injury. However, the effects of sorafenib on IR-induced liver injury in NASH rats had never been explored. In this study, we investigated the possible contributions of the Raf/MEK/ERK, p38MAPK, and Rho-kinase pathways to sorafenib-related effects on IR injury in NASH livers.

MATERIALS AND METHODS

IR Injury Model

To establish the IR liver model, the abdomen of each rat was opened, and the left hepatic artery and portal vein were occluded with a nontraumatic microvascular clip to induce ischemia of the left lateral and median lobes (approximately 70% of the total liver volume). After 90 minutes, the hepatic blood flow was restored, the abdomen was closed, and the rat was left to recover spontaneously. This model of nonlethal, partial IR injury avoided splanchnic congestion and thus any confounding effects resulting from bowel ischemia and hemodynamic disturbances. All procedures were approved by the laboratory animal care and use committee of Yang Ming Medical University and were conducted in accordance with Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication 86-23, revised 1985).

Animals

Male Sprague-Dawley rats (300-350 mg) were fed a methionine and choline–deficient diet for 12 weeks to induce NASH in their livers as reported elsewhere.1, 2

The IR liver model was established in the rats under intraperitoneal phenobarbital anesthesia (3.5 mL/kg) as described previously.2

Treatment and Sampling

The multikinase inhibitor sorafenib (200 mg) was purchased as Nexavar from Bayer Pharmaceuticals (West Haven, CT). After the removal of the outer coat, the pills were ground in a tissue mill in a manner similar to that of previous studies.7, 14 The resulting powder was mixed with 0.9% sodium chloride (normal saline) and was applied by gavage as previously reported.7, 14 In a preliminary dose study, different doses (10, 30, and 50 mg/kg/day) of sorafenib and the vehicle (0.9% sodium chloride) were given by oral gavage to NASH rats (n = 4 per dose) 24 hours before further IR study. This schedule was selected on the basis of the long half-life of sorafenib (24-48 hours).7, 14 In comparison with the vehicle-administered group, the abnormal liver function and increases in the necrotic index and leukocyte infiltration induced by IR were not significantly suppressed in NASH rats receiving sorafenib at 10 or 30 mg/kg before IR injury. In NASH rats that were acutely administered sorafenib at 50 mg/kg, the abnormal liver function and increases in the necrotic index and leukocyte infiltration induced by IR were significantly decreased in comparison with the vehicle-administered group. Therefore, this dose of sorafenib (50 mg/kg) or the vehicle was given to NASH rats by oral gavage 24 hours before further studies.

Fasudil [1-(5-isoquinolinesulfonyl)-homopiperazine] is an orally available Rho-kinase inhibitor that is well tolerated by human beings without any severe adverse reactions.12, 13 In a preliminary study, fasudil at 10 mg/kg was administered by intraperitoneal injection on the basis of real clinical applications and previous studies.12, 13 In our NASH rats, different durations of an intraperitoneal fasudil treatment (3, 5, or 7 days at 10 mg/kg/day before IR injury) were tested to suppress the IR-induced up-regulation of hepatic Rho-kinase protein expression. One group of NASH rats with IR injury that received saline was used as a control group for comparison (IR NASH rats). Studies showed that only 7 days of the intraperitoneal fasudil pretreatment at 10 mg/kg/day effectively suppressed the IR-induced up-regulation of hepatic Rho-kinase in NASH rats treated with fasudil versus IR NASH rats [p-Rho-kinase (where p indicates the phosphorylated form)/glyceraldehyde 3-phosphate dehydrogenase (GAPDH): 1.2132 ± 0.7452 versus 1.9821 ± 0.3459, P = 0.032].

Experimental Groups

Before the experiments, the NASH rats and the fasudil-pretreated NASH rats were assigned to the following 3 IR study groups (n = 8 per group):

  • 1Vehicle+IR group: NASH rats were acutely administered the vehicle by oral gavage 24 hours before ischemia.
  • 2Sorafenib+IR group: NASH rats were acutely administered sorafenib (50 mg/kg) by oral gavage 24 hours before ischemia.
  • 3Fasudil-sorafenib+IR group: NASH rats were pretreated for 7 days with fasudil (10 mg/kg) and then were acutely administered sorafenib (50 mg/kg) by oral gavage 24 hours before ischemia.

The hemodynamic parameters of all rats (n = 8 per group) were continuously monitored. Under general anesthesia, the rats were killed at different time points for the collection of serum and liver samples: before acute sorafenib or vehicle administration (basal time); before ischemia; and 1, 3, or 5 hours after reperfusion. In another set of rats, the hepatic microcirculation was evaluated.

Hemodynamic Parameters

In the second set of rats, the mean arterial pressure (MAP) was measured in accordance with a previous study.15 The hepatic tissue blood flow (HBF) was measured with a laser Doppler blood flowmeter (model LD 5000, Transonic Systemic). Each data point represented the mean of 5 consecutive measurements.

Measurements of Various Serum and Hepatic Cytokines and Chemokines

Blood was withdrawn from the inferior vena cava for the analysis of alanine aminotransferase (ALT) and cytokine/chemokine levels. Then, the serum and hepatic levels of tumor necrosis factor α (TNF-α), KC, MCP1, MIP2, granulocyte-monocyte colony-stimulating factor (GM-CSF), IL-1β, IL-6, and IL-8 were measured with individual enzyme-linked immunosorbent assay kits purchased from eBioscience (Bender MedSystems GmbH, Vienna, Austria).

Histological Staining for Necrosis Index Counting

Formalin-fixed liver samples were stained with hematoxylin-eosin (H-E) so that we could compute the area fraction of the necrotic parenchyma with an overlaid 1-point grid in 100 randomly sampled fields, and they were evaluated with Image-Pro Plus 6.1 (Media Cybernetics, Silver Spring, MD).

Immunostaining for Hepatic Apoptotic Bodies and Leukocyte Infiltration

Apoptotic bodies were measured with the ApopTag peroxidase in situ apoptosis detection kit (Millipore, Billerica, MA). Additionally, CD45 immunostaining was performed with a rabbit anti-mouse polyclonal antibody. Then, CD45-stained cells were counted in 20 nonconsecutive, randomly chosen histological fields at a magnification of ×100 (5 slides per animal). The results were expressed as the number of leukocytes per high-power field (HPF; ×10).

Direct Measurement of Caspase-3, Caspase-7, Caspase-8, and Caspase-9 Activity and DNA Fragmentation

Hepatic caspase-3, caspase-7, caspase-8, and caspase-9 activity was measured with caspase-3, caspase-7, caspase-8, and caspase-9 colorimetric assay kits (Axxora, Lörrach, Germany). For the DNA fragmentation analysis, a cell death detection enzyme-linked immunosorbent assay kit (Boehringer Mannheim, Indianapolis, IN) was used.5 In DNA fragmentation analysis, the kinetics of product generation (Vmax) is a measure of DNA fragmentation. The Vmax values obtained for the untreated controls (the vehicle group; 100%) were compared with the values obtained for the treated groups. The assay allowed the specific quantification of histone-associated DNA fragments (mononucleosomes and oligonucleosomes) induced by caspase-activated endonucleases during apoptosis as well as DNA fragmentation during necrosis in the cytoplasmic fraction of cell lysates.

Measurement of Various Protein and Messenger RNA (mRNA) Levels

The protein expression of p-Raf1, p-MEK 1/2 (Ser217/Ser221), p-ERK1/2 (Tyr202/Tyr204), p-Rho-kinase (Thr558), p-Akt (Ser477), p-p38MAPK (Thr180/Tyr182), vascular cell adhesion molecule 1 (VCAM1), intercellular adhesion molecule 1 (ICAM1), and GAPDH was measured with a western blotting analysis.

Meanwhile, the mRNA levels of B cell lymphoma 2–associated death promoter (Bad), B cell lymphoma 2–associated X protein (Bax), B cell lymphoma extra large (Bcl-xL), B cell lymphoma 2 (Bcl2), and β-actin (the control) were measured by quantitative real-time reverse-transcription polymerase chain reaction with appropriate primers [Bad, 5-CGGAGGATGAGTGACGAGTT-3 (sense primer) and 5-GATGTGGAGCGAAGGTCACT-3 (antisense primer); Bax, 5-TCCCCCCGAGAGGTCTTTT-3 (sense primer) and 5-CGGCCCCAGTTGAAGTTG-3 (antisense primer); Bcl-xL, 5-TCCTTGTCTACGCTTTCCACG-3 (sense primer) and 5-GGTCGCATTTGTGGCCTTT-3 (antisense primer); Bcl2, 5-CATGTGTGTGGAGAGCGTCAA-3 (sense primer) and 5-GCCGGTTCAGGTATCAGTCA-3 (antisense primer); and β-actin, 5-TGGAGAAGAGTCATGAGCTGCCT-3 (sense primer) and 5-GTGCCACCAGACAGCACTGTGT-3 (antisense primer)] on an ABI-Prism 7000 sequence detection system (Applied Biosystems, Tokyo, Japan).

Analysis of Microcirculatory Dysfunction by Intravital Fluorescence Microscopy

The parameters measured to represent hepatic microcirculatory dysfunction included (1) the sinusoidal perfusion rate (ie, the percentage of perfused sinusoids among all observed sinusoids), (2) the sinusoidal diameter [ie, the average diameter of 8-14 sinusoids in the midzonal area of each acinus (%)], (3) the postsinusoidal venule diameter [ie, the diameter at an identical position in each postsinusoidal venule (%)], (4) the number of sticky leukocytes in sinusoids [ie, the number of stained cells located within sinusoids and not moving during an observation period of 30 seconds (cells/lobule)], and (5) the number of adherent leukocytes in postsinusoidal venules [ie, the number of stained cells located in the postsinusoidal venule and attached to the endothelial lining during an observation period of 30 seconds (cells/mm2 of endovascular section)]. The diameters of sinusoids and postsinusoidal venules were expressed as percentages with respect to the baseline values before ischemia.16 The heart rate, MAP, and portal vein pressure were monitored continuously.

Materials

Antibodies against p-Raf1, p-MEK1/2 (Ser217/Ser221), p-ERK1/2 (Tyr202/Tyr204), p-Akt (Ser477), and GAPDH were purchased from Cell Signaling Technology (Danvers, MA). Rabbit anti–p-Rho-kinase Thr558 antibodies, p-p38 MAP (Thr180/Tyr182), VCAM1, and ICAM1 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Primers for Bad, Bax, Bcl-xL, Bcl2, and β-actin were purchased from Applied Biosystems. All other reagents were obtained from Sigma (St. Louis, MO).

Statistical Analysis

All values are presented as means and standard deviations. Intergroup differences were analyzed with a nonparametric Kruskal-Wallis analysis of variance rank test. When the differences were significant, a Student-Newman-Keuls test was applied. P < 0.05 was considered to be significant.

Abbreviations:

ALT, alanine aminotransferase; Bad, B cell lymphoma 2–associated death promoter; Bax, B cell lymphoma 2–associated X protein; Bcl2, B cell lymphoma 2; Bcl-xL, B cell lymphoma extra large; ERK, extracellular signal-regulated kinase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GM-CSF, granulocyte-monocyte colony-stimulating factor; HBF, hepatic tissue blood flow; H-E, hematoxylin-eosin; HPF, high-power field; ICAM1, intercellular adhesion molecule 1; IL, interleukin; IR, ischemia/reperfusion; KC, keratinocyte chemoattractant; MAP, mean arterial pressure; MAPK, mitogen-activated protein kinase, MCP1, monocyte chemoattractant protein 1; MEK, mitogen-activated protein–extracellular signal-regulated kinase kinase; MIP2, macrophage inflammatory protein 2; mRNA, messenger RNA; NASH, nonalcoholic steatohepatitis; p, phosphorylated form; p38MAPK, p38 mitogen-activated protein kinase; TNF-α, tumor necrosis factor α; VCAM1, vascular cell adhesion molecule 1; Vmax, kinetics of product generation.

RESULTS

Sorafenib Prevents Decreases in MAP and HBF Induced by IR Injury

In vehicle+IR rats with NASH, MAP significantly decreased from 118 to 80 mm Hg (67%) 1 hour after reperfusion (P < 0.05); it recovered to approximately 83% (98 mm Hg) of the basal value 3 hours after reperfusion. In sorafenib+IR rats with NASH, the hypotension observed in the initial phase [110 ⇒ 96 mm Hg (87%)] was antagonized by the acute administration of sorafenib (Fig. 1A). Similarly, HBF was markedly decreased 1 hour after reperfusion in vehicle+IR rats with NASH [21 ⇒ 8 AU (38%)], and recovery was observed 3 hours after reperfusion [13 AU (61%); Fig. 1B]. However, the IR-induced decrease in HBF was prevented by the acute administration of sorafenib in the sorafenib+IR rats with NASH [19 ⇒ 14 AU (73%)]. In all the animals studied, a single dose of sorafenib was well tolerated with no signs of toxicity, adverse effects (reduced body weight, diarrhea, and hemorrhaging), or drug-induced toxicity.

Figure 1.

Serial changes in (A) MAP, (B) HBF, and (C) serum ALT levels after IR. *P < 0.05 versus the previous time point, #P < 0.05 versus the vehicle+IR group, and †P < 0.05 versus the sorafenib+IR group.

Although statistical significance was not reached, chronic pretreatment with the Rho-kinase inhibitor fasudil resulted in a trend of abolishing sorafenib-related effects on MAP and HBF in fasudil-sorafenib+IR rats with NASH.

Sorafenib Inhibits Serum and Hepatic Cytokine Release and Hepatic Inflammation During Acute Injury Induced by IR

Cytokines were substantially increased in blood and liver samples collected from all NASH rats 5 hours after reperfusion (Table 1). Notably, only increases in serum and hepatic levels of MCP2, KC, and GM-CSF were significantly inhibited by the acute administration of sorafenib (Table 1). Meanwhile, the sorafenib-related inhibition of these circulating and hepatic cytokines was abolished by chronic pretreatment with fasudil.

Table 1. Levels of Various Serum and Hepatic Cytokines and Chemokines
 Baseline Before Ischemia5 Hours After Reperfusion
Vehicle+IRSorafenib+IRFasudil-Sorafenib+IRVehicle+IRSorafenib+IRFasudil-Sorafenib+IR
  • NOTE: The data are presented as means and standard deviations.

  • *

    P < 0.01 versus the baseline values in the corresponding group.

  • P < 0.05 versus the vehicle+IR group.

  • P < 0.05 versus the sorafenib+IR group.

Serum MCP1 (pg/mL)28 ± 726 ± 529 ± 121007 ± 197 (P = 0.001)*911 ± 29 (P = 0.006)*928 ± 179 (P < 0.001)*
Serum MIP2 (pg/mL)14 ± 310 ± 2 (P = 0.03)12 ± 35600 ± 554 (P = 0.002)*2912 ± 179 (P = 0.003, P = 0.04)*4012 ± 372 (P = 0.003, P = 0.02)*
Serum KC (pg/mL)212 ± 6180 ± 12 (P = 0.02)208 ± 28 (P = 0.02)1520 ± 408 (P = 0.006)*638 ± 140 (P = 0.006, P = 0.04)*997 ± 249 (P = 0.003, P = 0.04)*
Serum GM-CSF (pg/mL)98 ± 1354 ± 9 (P = 0.04)84 ± 11 (P = 0.03)702 ± 140 (P = 0.009)*420 ± 54 (P = 0.008, P = 0.02)*589 ± 147 (P = 0.001, P = 0.04)*
Serum IL-6 (pg/mL)244 ± 17196 ± 31226 ± 291750 ± 179 (P = 0.004)*1362 ± 574 (P = 0.007)*1200 ± 372 (P = 0.004)*
Serum IL-10 (pg/mL)58 ± 846 ± 949 ± 13392 ± 21 (P = 0.003)*299 ± 51 (P = 0.004)*300 ± 70 (P = 0.002)*
Serum IL-1β (pg/mL)32 ± 428 ± 322 ± 9190 ± 65 (P = 0.007)*110 ± 25 (P = 0.002)*101 ± 19 (P < 0.001)*
Serum TNF-α (pg/mL)36 ± 1130 ± 434 ± 9260 ± 74 (P = 0.005)*230 ± 27 (P = 0.005)*210 ± 23 (P = 0.003)*
Hepatic MIP2 (pg/g)8 ± 1.65 ± 1 (P = 0.02)7 ± 1.75100 ± 532 (P = 0.002)*1700 ± 168 (P = 0.003, P = 0.04)*4102 ± 352 (P = 0.005, P = 0.03)*
Hepatic KC (pg/g)215 ± 6082 ± 12 (P = 0.07)206 ± 27 (P = 0.03)1900 ± 974 (P = 0.004)*789 ± 343 (P = 0.002, P = 0.03)*1501 ± 580 (P = 0.002, P = 0.04)*
Hepatic GM-CSF (pg/g)100 ± 1355 ± 8.8 (P = 0.03)86 ± 11.2 (P = 0.04)700 ± 119 (P = 0.002)*480 ± 62 (P = 0.004, P = 0.02)*570 ± 125 (P < 0.001, P = 0.02)*

Additionally, the levels of serum ALT, a marker of hepatic inflammation, were similar for vehicle+IR, sorafenib+IR, and fasudil-sorafenib+IR rats with NASH at the baseline (pre-ischemia stage; Fig. 1C). One hour after reperfusion, the serum ALT level was elevated up to 1349 U/L in vehicle+IR rats with NASH, and it peaked 5 hours after reperfusion. As expected, the IR-related elevation of serum ALT levels was clearly milder in sorafenib+IR rats with NASH versus vehicle+IR rats with NASH. Nonetheless, the sorafenib-related protection of NASH rat livers against damage was abolished by chronic pretreatment with the Rho-kinase inhibitor fasudil.

Sorafenib Attenuates Hepatic Leukocyte Infiltration and Necrosis During Acute Liver Injury Induced by IR

Hepatic leukocyte infiltration progressively increased in all NASH rats subjected to IR injury. A protective effect of sorafenib was observed as a lower increase in the leukocyte count and as decreased protein expression of CD45, VCAM1, and ICAM1 in the livers of sorafenib+IR rats with NASH (Figs. 2A,B, 3D,E, and 4E,F). Furthermore, the sorafenib-related inhibition of hepatic leukocyte infiltration and down-regulation of hepatic CD45, VCAM1, and ICAM1 proteins were abolished by chronic pretreatment with fasudil.

Figure 2.

Representative pictures of (A) CD45 and (C) H-E staining and serial changes in (B) the number of leukocytes and (D) the necrotic index after IR in the livers of all NASH rats. *P < 0.05 versus the previous time point, #P < 0.05 versus the vehicle+IR group, and †P < 0.05 versus the sorafenib+IR group.

Figure 3.

Percent changes in (A) the sinusoid perfusion rate, (B) the sinusoid diameter, and (C) the presinusoidal venule diameter of NASH rat livers; (D) number of sticky leukocytes in the sinusoids of NASH rat livers; and (E) number of sticky leukocytes in the presinusoidal venules of NASH rat livers (n = 8 per group). *P < 0.05 versus the previous time point, #P < 0.05 versus the vehicle+IR group, and †P < 0.05 versus the sorafenib+IR group.

Figure 4.

Representative images of the hepatic protein expression of (A) p-Rho-kinase, (B) p-Raf1, (C) p-MEK1/2, (D) p-ERK1/2, (E) VCAM1, and (F) ICAM1 in all NASH rats. Lanes 1, 2, 3, and 4 represent the control, vehicle+IR, sorafenib+IR, and fasudil-sorafenib+IR groups, respectively. *P < 0.05 versus the vehicle group, #P < 0.05 versus the vehicle+IR group, and †P < 0.05 versus the sorafenib+IR group.

IR also markedly induced necrosis in NASH rat livers, which peaked at 3 hours and progressively returned 5 hours after reperfusion. Consistently, IR-induced hepatic necrosis was significantly reversed by sorafenib in sorafenib+IR rats with NASH. In addition, the sorafenib-related improvement of liver necrosis was diminished by chronic pretreatment with fasudil in fasudil-sorafenib+IR rats with NASH (Fig. 2C,D).

Sorafenib Suppresses Hepatic Apoptosis in Acute Liver Injury Induced by IR

IR markedly increased hepatic caspase-3, caspase-7, caspase-8, and caspase-9 activity in all NASH rat livers, and this increase peaked 5 hours after reperfusion. Significantly, the increased hepatic caspase-3 and caspase-9 activity during perfusion was reversed by sorafenib in the livers of sorafenib+IR rats with NASH 5 hours after reperfusion. Furthermore, the sorafenib-related suppression of increased hepatic caspase-3 and caspase-9 activity was prevented by chronic pretreatment with fasudil (Table 2). According to immunochemical staining, hepatic apoptotic bodies in NASH rat livers were markedly up-regulated by IR and attenuated by the acute administration of sorafenib. Consistently, chronic pretreatment with the Rho-kinase inhibitor fasudil abolished the sorafenib-related inhibition of the up-regulation of hepatic apoptosis in NASH rat livers (Fig. 5A).

Figure 5.

(A) Representative pictures of apoptotic bodies with immunohistochemical staining (with arrows indicating areas in the liver positively stained for apoptotic bodies) and mRNA levels of (B) proapoptotic markers (Bad and Bax) and (C) antiapoptotic markers (Bcl-xL and Bcl2) in the livers of all NASH rats. Lanes 1, 2, and 3 represent the vehicle+IR, sorafenib+IR, and fasudil-sorafenib+IR groups, respectively. *P < 0.01 versus the vehicle+IR group, and #P < 0.05 versus the sorafenib+IR group.

Table 2. Hepatic Apoptotic Activity in NASH Rats
 Baseline Before Ischemia5 Hours After Reperfusion
Vehicle+IRSorafenib+IRFasudil- Sorafenib+IRVehicle+IRSorafenib+IRChange Versus Vehicle+IR GroupFasudil- Sorafenib+IRChange Versus Vehicle+IR Group
  • NOTE: The data are presented as means and standard deviations.

  • *

    P < 0.05 versus the vehicle+IR group.

  • P < 0.01 versus the baseline values.

  • P < 0.05 versus the sorafenib+IR group.

Caspase-3 activity (absorbance/μg of protein)0.043 ± 0.0060.012 ± 0.003 (P = 0.02)*0.037 ± 0.0080.11 ± 0.012 (P = 0.004)0.05 ± 0.002 (P = 0.004, P = 0.02)*−55% ± 23%0.09 ± 0.0029 (P = 0.002, P = 0.04)−34% ± 31%
Caspase-7 activity (absorbance/μg of protein)0.053 ± 0.0040.051 ± 0.0020.049 ± 0.0080.097 ± 0.009 (P = 0.006)0.079 ± 0.003 (P = 0.003)−3% ± 2.2%0.082 ± 0.007 (P = 0.003)−9% ± 1.7%
Caspase-8 activity (absorbance/μg of protein)0.008 ± 0.0010.011 ± 0.0030.009 ± 0.00030.056 ± 0.027 (P = 0.002)0.052 ± 0.023 (P = 0.005)−8% ± 1.3%0.053 ± 0.033 (P = 0.004)−6% ± 2.1%
Caspase-9 activity (absorbance/μg of protein)0.014 ± 0.0010.009 ± 0.001 (P = 0.04)*0.011 ± 0.00780.08 ± 0.027 (P = 0.005)0.06 ± 0.015 (P = 0.002, P = 0.01)*−25% ± 8%0.074 ± 0.023 (P = 0.003, P = 0.04)−5% ± 2.1%
DNA fragmentation (Vmax %)78 ± 1570 ± 271 ± 9602 ± 21 (P = 0.003)240 ± 3 (P = 0.007, P = 0.03)*−71% ± 8%423 ± 57 (P = 0.002, P = 0.01)−34% ± 3.1%

In comparison with vehicle+IR rats with NASH, significant down-regulation of the mRNA levels of hepatic proapoptotic markers (Bad/Bax) and up-regulation of the mRNA levels of antiapoptotic markers (Bcl-xL/Bcl2) were noted in sorafenib+IR rats with NASH. Moreover, the sorafenib-related down-regulation of the mRNA levels of hepatic proapoptotic markers (Bad/Bax) and up-regulation of the mRNA levels of antiapoptotic markers (Bcl-xL/Bcl2) were abolished by chronic pretreatment with the Rho-kinase inhibitor fasudil in the livers of fasudil-sorafenib+IR rats with NASH (Fig. 5B,C).

In NASH rat livers, DNA fragmentation was substantially increased during reperfusion with peak values 771% greater than the baseline values 5 hours after reperfusion (602% ± 21% versus 78% ± 15%; Table 2). In sorafenib+IR rats with NASH, hepatic DNA fragmentation was significantly suppressed. Similarly, the sorafenib-related inhibition of hepatic DNA fragmentation was attenuated by the inhibition of hepatic Rho-kinase activity by chronic pretreatment with fasudil. Notably, the trend in hepatic DNA fragmentation was paralleled by corresponding changes in the hepatic apoptotic activity, leukocyte infiltration, and necrotic index of fasudil-sorafenib+IR rats with NASH (Table 2 and Figs. 2B,D, 3D,E, and 5A).

Sorafenib Improves Hepatic Microcirculatory Dysfunction in NASH Rat Livers

As shown in Fig. 3, markedly decreased sinusoidal perfusion rates and sinusoidal/postsinusoidal venule diameters were observed in all NASH rat livers 1 hour after reperfusion (in comparison with the baseline), and they slightly recovered within 5 hours versus the 1-hour values. Meanwhile, the acute administration of sorafenib significantly attenuated the postischemic perfusion failure and vasoconstriction of sinusoids and postsinusoidal venules. Nonetheless, the sorafenib-related effects on sinusoidal perfusion and sinusoidal/postsinusoidal venule diameters were partially abolished by chronic pretreatment with fasudil in the livers of fasudil-sorafenib+IR rats with NASH. Notably, the time points for liver perfusion reduction and recovery (Fig. 1B) paralleled the changes in sinusoidal perfusion and in the diameters of sinusoids and postsinusoidal venules (Fig. 3A-C) in all NASH rat livers. On the other hand, leukocyte-endothelial interactions, which were evaluated from the numbers of sticky and adherent leukocytes in sinusoids and postsinusoidal venules, were markedly enhanced by ischemia and peaked 1 hour after reperfusion in all NASH rat livers. In the livers of sorafenib+IR rats with NASH, leukocyte-endothelial interactions were significantly reduced in comparison with vehicle+IR rats with NASH. Moreover, chronic fasudil pretreatment significantly prevented the beneficial effects of sorafenib on IR-stimulated leukocyte-endothelial interactions in fasudil-sorafenib+IR rats with NASH (Fig. 3).

Mechanisms of the Protective Effect of Sorafenib in NASH Rat Livers: Down-Regulation of the Hepatic Rho/Akt Kinase and Raf/MEK/ERK Pathways

In comparison with control NASH rats without IR injury, the expression of hepatic p-Rho-kinase, p-Raf1, p-MEK1/2, and p-ERK1/2 proteins was markedly up-regulated by IR injury (Fig. 4A-D). Furthermore, the acute administration of sorafenib significantly suppressed the IR injury–induced up-regulation of p-Rho-kinase proteins as well as p-Raf1, p-MEK1/2, and p-ERK1/2 proteins in the livers of sorafenib+IR rats with NASH. Interestingly, the presuppression of hepatic p-Rho-kinase protein expression by chronic pretreatment with fasudil significantly blocked the sorafenib-related down-regulation of p-Raf1, p-MEK1/2, and p-ERK1/2 protein expression in the livers of fasudil-sorafenib+IR rats with NASH. Significant up-regulation of the hepatic protein expression of p-Akt (p-Akt/GAPDH: 1.13645 ± 0.0472 versus 0.7892 ± 0.08213, P = 0.0215) and p-p38MAPK (p-p38MAPK/GAPDH: 0.9972 ± 0.0615 versus 0.5481 ± 0.0711, P = 0.0132) was noted in vehicle+IR rats with NASH versus control NASH rats without IR injury. In comparison with vehicle+IR rats with NASH, the hepatic protein expression of p-Akt and p-p38MAPK was not modified in either sorafenib+IR rats with NASH (p-Akt/GAPDH: 1.07945 ± 0.1213 versus 1.13645 ± 0.0472, P = 0.89; p-p38MAPK/GADPH: 0.9621 ± 0.1001 versus 0.9972 ± 0.0615, P = 0.713) or fasudil-sorafenib+IR rats with NASH (p-Akt/GAPDH: 1.231 ± 0.0923 versus 1.13645 ± 0.0472, P = 1.018; p-p38MAPK/GADPH: 1.099 ± 0.0932 versus 0.9972 ± 0.0615, P = 0.876).

DISCUSSION

This study has demonstrated that sorafenib is able to reverse the hypotensive effects of IR liver injury in NASH rats. Notably, a temporal relationship between the improvement in MAP occurring after reperfusion and HBF suggests that the preservation of systemic hemodynamics favors better revascularization of the liver and leads to the faster recovery of tissue perfusion in sorafenib-administered NASH rats.

IR liver injury is characterized by hepatic inflammation and leukocyte recruitment. Adhesion molecules, including VCAM1 and ICAM1, are up-regulated, and cytokine release stimulates hepatic inflammation and leukocyte recruitment in IR rat livers.6 Consequently, hepatic microcirculatory dysfunction develops because of sinusoidal/postsinusoidal venule occlusion and hypoperfusion.4, 6, 17 Hypoxia and hypoperfusion induce Rho-kinase activation in the pulmonary arteries and kidneys of rats.18, 19 In particular, Rho-kinase signaling is involved in leukocyte activation and migration in septic livers and IR-injured kidneys.9, 19 Moreover, Rho-kinase inhibition significantly prevents hepatic leukocyte infiltration and inflammation in rats with acute liver injury.20 In our NASH rats with IR injury, hepatic inflammation and microcirculatory dysfunction were accompanied by leukocyte recruitment, cytokine release, and the up-regulation of Rho-kinase and adhesion molecules (VCAM1 and ICAM1). Subsequently, the aforementioned IR injury–stimulated effects were markedly suppressed by acute sorafenib administration in the livers of sorafenib+IR rats with NASH. Additionally, chronic pretreatment with the Rho-kinase inhibitor fasudil abolished sorafenib-related protective effects. These results suggest that sorafenib protects NASH rats from IR liver injury through the Rho-kinase–dependent inhibition of adhesion molecule expression, cytokine release, leukocyte infiltration, hepatic inflammation, and microcirculatory dysfunction.

TNF-α has been implicated in the pathogenesis of hepatic IR injury; this occurs through the stimulation of the expression of adhesion molecules (VCAM1 and ICAM1) and the release of leukocyte-attracting chemokines.21, 22 However, the IR-stimulated elevation of serum and hepatic TNF-α levels was not modified by acute sorafenib administration in our NASH rats. This observation indicates that sorafenib-related protective effects against IR liver injury, including the down-regulation of VCAM1 and ICAM1 and the inhibition of leukocyte infiltration, are mediated by mechanisms other than the TNF-α signal pathway.

In the liver, KC and GM-CSF are involved in increasing leukocyte activation, rolling, adhesion, and recruitment.23 Moreover, potent chemokines and cytokines, including MCP1, MIP2, IL-1β, IL-6, and IL-10, have been positively correlated with the degree of hepatic leukocyte recruitment in rats with acute IR injury and chronic liver injury.22-24 Notably, the levels of MCP1, MIP2, KC, GM-CSF, IL-6, IL-10, and IL-1β were significantly increased after IR liver injury in our NASH rats. However, only decreases in serum and hepatic MIP2, KC, and GM-CSF levels were found to be associated with the inhibition of hepatic leukocyte infiltration and VCAM1/ICAM1 expression when the group acutely administered sorafenib was compared to the vehicle-administered group. These results support the hypothesis that MIP2, KC, and GM-CSF are significant factors associated with reducing leukocyte infiltration when NASH rats with IR liver injury are treated with sorafenib.

IR liver injury is characterized by increased hepatic necrosis.5 Mechanistically, hepatic necrosis is a result of inflammation and microcirculatory dysfunction. The present study suggests that both inflammation and microcirculatory dysfunction are mediated by Rho-kinase–dependent mechanisms after IR liver injury in rats with NASH. When the results are considered together, it is reasonable to conclude that the sorafenib-related antinecrosis, inflammation, and microcirculatory dysfunction effects were simultaneously attenuated by chronic pretreatment with the specific Rho-kinase inhibitor fasudil.

Increased hepatic apoptosis, which is positively correlated with the degree of hepatic necrosis, has been documented in NASH.25 Moreover, hepatic apoptosis is initiated shortly after ischemia and is amplified by reperfusion injury.5, 25 Notably, sorafenib significantly inhibited IR injury–induced apoptotic cascades (including caspase-3, caspase-9, Bad, and Bax) and DNA fragmentation in our NASH rats. These sorafenib-related antiapoptotic effects were accompanied by the up-regulation of antiapoptotic markers (Bcl-xL and Bcl2) and a reduction in hepatic necrosis.

The Rho/Rho-kinase pathway, which can be down-regulated by the acute administration of sorafenib, is known to be involved in the modulation of hepatic apoptosis.7, 13 Furthermore, Rho-kinase inhibition has been shown to attenuate hepatic apoptosis.20 In our NASH rats with IR injury, the sorafenib-related protective effects were Rho/Rho-kinase–dependent and were associated with antiapoptotic effects in livers. Mechanistically, the Akt pathway has been shown to participate in the regulation of cell apoptosis.26 The activation of the Akt pathway seems to be involved in the Rho-kinase inhibition–related attenuation of apoptosis after acute liver injury.20 Modulation of the Akt pathway probably contributes to the sorafenib-related suppression of hepatic apoptosis during IR injury. In our study, the increased hepatic apoptosis was accompanied by the up-regulation of hepatic p-Akt and Rho/Rho-kinase proteins in vehicle+IR rats with NASH. Nonetheless, we found that the suppression of hepatic apoptosis by sorafenib was not accompanied by Akt pathway modification in the livers of NASH rats with IR injury. These findings indicate that a signal pathway other than the Akt pathway mediated the sorafenib-related Rho-kinase–dependent antiapoptotic effects in the livers of our NASH rats.

Signaling cascades, including the Rho-kinase, Raf/MEK/ERK, and p38MAPK pathways, have been implicated in the IR liver injury process.10, 13, 27, 28 In the livers of our NASH rats, Rho-kinase, Raf1, MEK1/2, ERK1/2, and p38MAPK protein expression was measured to clarify the upstream and downstream signal cascades related to sorafenib effects. Simultaneously, IR liver injury stimulated hepatic Rho-kinase, p38MAPK, Raf1, MEK1/2, and ERK1/2 kinase phosphorylation in our NASH rats. Furthermore, the acute administration of sorafenib markedly attenuated the IR-stimulated up-regulation of Rho-kinase, Raf1, MEK1/2, and ERK1/2 kinase phosphorylation in the livers of our NASH rats. Nonetheless, the phosphorylation of p38MAPK was not modified by acute sorafenib administration. As for the Raf/MEK/ERK pathway, MEK1/2 MAPK is known to be the direct upstream kinase of ERK1/2 MAPK, which is activated by upstream kinase-Raf1 phosphorylation.10, 13, 27, 28 Before we consider the clinical implications, it is important to clarify whether Raf1 or Rho-kinase is the most upstream cascade associated with the effects of sorafenib. Recently, it has been reported that acute Rho-kinase inhibition improves the survival rate of rats with NASH after IR liver injury.13 In the livers of our NASH rats with IR injury, chronic pretreatment with the Rho-kinase inhibitor fasudil markedly attenuated the sorafenib-related modification of hepatic Raf1, MEK1/2, and ERK1/2 phosphorylation. Taken together, our results suggest that Rho-kinase down-regulation is the earliest target for hepatic IR injury–induced activation of the Raf/MEK/ERK pathway in sorafenib-administered NASH rats.

Multiple mechanisms, including inflammatory mediator release and multikinase activation, seem to be involved in the impaired tolerance of NASH livers for IR injury. Obviously, specific pharmacological strategies that simultaneously suppress multiple IR injury–stimulated mechanisms are needed to effectively protect NASH livers during transplantation. We have demonstrated that the multikinase inhibitor sorafenib protects NASH livers against IR injury by interfering with hepatic inflammation, necrotic, and apoptotic responses that cause leukocyte-dependent hepatic microcirculatory dysfunction (Fig. 6). Mechanistically, the hepatoprotective effects of sorafenib seem to occur via the suppression of the Rho-kinase–dependent Raf/MEK/ERK pathway, which is up-regulated during IR in NASH rat livers.

Figure 6.

Schematic representation of sorafenib's protection of NASH rats against IR liver injury.

In conclusion, our study suggests that multiple mechanisms are manipulated by sorafenib and that this drug could be potentially useful for helping to prevent IR injury in NASH livers. However, this is the first study to discover the protective effects of sorafenib in NASH rats with IR liver injury. Thus, a systematic evaluation of possible harmful effects is mandatory before sorafenib is applied to protect NASH animals from IR liver injury during liver transplantation. In clinical practice, we hope that our study can push the limitations of NASH liver grafts to increase the organ supply and decrease the incidence of NASH graft failure in the future.

Ancillary