TOXICOLOGICAL SCIENCES, 143(2), 2015, 430–440
doi: 10.1093/toxsci/kfu244
Advance Access Publication Date: November 27, 2014
Cyclooxygenase-1 Serves a Vital Hepato-Protective
Function in Chemically Induced Acute Liver Injury
Jia Xiao*,†, Emily C. Liong†, Hai Huang†, Wing On Tse†, Kam Shing Lau†,
Jingfei Pan‡, Amin A. Nanji§, Man Lung Fung‡, Feiyue Xing*, and
George L. Tipoe†,1
*Department of Immunobiology, Institute of Tissue Transplantation and Immunology, Jinan University,
Guangzhou, 510632 China, †Department of Anatomy and ‡Department of Physiology, Li Ka Shing Faculty of
Medicine, The University of Hong Kong, Hong Kong, China and §Department of Pathology and Laboratory
Medicine, Dalhousie University School of Medicine, Halifax, B3H 4R2 Canada
1
To whom correspondence should be addressed. Fax: þ852 28170857. E-mail: [email protected].
ABSTRACT
Cyclooxygenase-1 (COX-1) is the constitutive form of the COX enzyme family, which produces bioactive lipids called
prostanoids. Although the role of COX-2 in liver diseases has been studied, little is known about the function of COX-1 in
liver injury. We aimed to investigate the role and mechanism of COX-1 in acute liver injury. Carbon tetrachloride (CCl4) was
administered to induce acute liver injury in wild-type or COX-1-deficient mice. Both genetic (partially or completely)
deletion of COX-1 expression and pharmacological inhibition of COX-1 activity in mice exacerbated acute liver injury
induced by CCl4, revealing the (1) histopathological changes and increased serum levels of aminotransferases; (2) oxidative
stress in the liver partly through the action of cytochrome P450 2E1-dependent pathway; (3) enhanced inflammatory and
chemoattractive responses with increased number of activated macrophages; and (4) increased apoptosis through both
intrinsic and extrinsic apoptotic pathways. These pathological changes were partly through the modulation of transcription
factor-dependent pathways (eg, NF-jB and C/EBP-a). Pre-treatment with prostaglandin E2 (PGE2) or 5-lipoxygenase (5-LO)
inhibitor in homozygous COX-1 knockout mice significantly ameliorated CCl4-induced hepatic injury. In addition, level of
hepato-protective molecules (eg, OSM and OSMR) and associated liver regeneration pathway were significantly inhibited by
the deficiency of COX-1 but restored by the addition of PGE2 or the inhibition of 5-LO. Furthermore, the alternative
arachidonic acid metabolism pathway of 5-LO, which induced additional inflammation in the liver, was activated in
response to the deficiency of COX-1. In conclusion, basal expression of COX-1 is essential for the protection of liver against
chemical-induced hepatotoxicity and required for hepatic homeostatic maintenance.
Key words: cyclooxygenase; carbon tetrachloride; prostaglandin; acute liver injury; 5-lipoxygenase
Acute liver injury is the appearance of liver cell damage that
suddenly occurs within a short period of time. Severe acute liver
injury, if not successfully controlled, may progress to acute liver
failure which is life-threatening, and with a high mortality rate
up to 80% (Gill and Sterling, 2001). The main consequence of
acute liver injury is cellular death induced by oxidative stress
and inflammation. There are two typical patterns of cellular
death occurring in acute liver injury namely, necrosis and apoptosis. The intrinsic and extrinsic pathways are the two main
pathways of apoptosis that have been described in acute liver
injury (Rutherford and Chung, 2008; Spencer and Sorger, 2011).
Carbon tetrachloride (CCl4) is a common chemical toxin that induces acute liver injury. In the liver, the metabolism of CCl4 begins with the production of the trichloromethyl free radical
C The Author 2014. Published by Oxford University Press on behalf of the Society of Toxicology.
V
All rights reserved. For Permissions, please e-mail: [email protected]
430
XIAO ET AL.
CCl3 through reactions of cytochrome P450 enzymes (primarily
P450 2E1) on the endoplasmic reticulum (Khan et al., 2012). Upon
stimulation of CCl4 or its derivatives, Kupffer cells are activated
to release tumor necrosis factor-alpha (TNF-a), nitric oxide (NO),
interleukin-1beta (IL-1b), IL-6, and IL-10. These molecules trigger both oxidative stress and inflammatory response in the liver
(Liu et al., 2010). Due to the close association among inflammation, cellular injury, and apoptosis, transcription factors such as
nuclear factor kappa B (NF-jB) and Ccaat-enhancer-binding proteins-alpha (C/EBP-a) are critical in mediating the upstream
pathological events and downstream transcriptional regulations (Chen et al., 2004).
Cyclooxygenases (COXs) are enzymes responsible for the formation of bioactive lipids called prostanoids, including prostaglandins (PGs), prostacyclin, and thromboxanes. Prostanoids
are involved in a spectrum of important biological processes
including gastric mucosa integrity, inflammatory responses, angiogenesis, and carcinogenesis (Cha and DuBois, 2007; MartinSanz et al., 2010; Wang and Dubois, 2006). Currently, there are 3
isoforms of COXs including COX-1, COX-2, and COX-1b, a splice
variant of COX-1 (Chandrasekharan et al., 2002). COX-1 is the
constitutive form of COX which continuously synthesizes prostanoids, whereas COX-2 is the inducible form which is activated
in response to a variety of cellular stresses, such as pro-inflammatory cytokines, growth factors, and hormones (Daikoku et al.,
2006; Harris et al., 2002). The role of COX-2 in hepatic pathological conditions has been extensively investigated in past decades. Continuous supply of bioactive lipids both from
arachidonic acid and other polyunsaturated fatty acids by COX2 in the liver prevents acute injury (Martin-Sanz et al., 2010). It is
also known that inhibition of COX-1 induces apoptosis in human hepatocellular carcinoma cells (Lampiasi et al., 2006) and
contributes to gastric mucosal apoptosis (Wu et al., 2009).
However, the role of COX-1 in liver injury remains largely unknown. In the current study, we showed that a basal expression
and downstream prostaglandin E2 (PGE2) production of COX-1
were vital for the hepatic defense against chemical-induced
acute liver injury. Genetic deletion and pharmacological
inhibition of COX-1 exacerbated CCl4 induced acute liver
injury. The arachidonic acid metabolism pathway (primarily
5-lipoxygenase [5-LO]) also contributed to the protective role of
COX-1.
MATERIALS AND METHODS
Materials. CCl4 was purchased from Tianjin Baishi Chemical
(Tianjin, China). COX-1 activity-specific blocker SC-560 (COX-1
IC50 ¼ 9 nM; COX-2 IC50 ¼ 6.3 lM) and synthetic PGE2 were from
Cayman Chemical (Ann Arbor, Michigan) and Sigma-Aldrich (St
Louis, Missouri), respectively. Specific 5-LO inhibitor, AA-861
was purchased from Sigma-Aldrich. Primary antibodies were
bought from following companies: b-actin, Sigma-Aldrich; COX1, TIMP1, ED1 (rat homolog of human CD68), and Bcl-2, Santa
Cruz Biotechnology (Santa Cruz, California); C/EBP-a, cleaved
caspase-3, cytochrome C (cyto C), Bak1, cleaved caspase-8, FasL,
phosphorylated inhibitor of NF-jB alpha (IjBa) at Ser32 (pIjBa),
IjBa, phosphorylated STAT3 at Tyr705 (pSTAT3), and STAT3,
Cell Signaling Technology (Danvers, Massachusetts); Cytochrome P450 2E1 (CYP2E1), Millipore (Billerica, Massachusetts);
Nitrotyrosine (NTR), Zymed (San Francisco, California).
32
P radionuclide and NF-jB gel-shift assay (electrophoretic
mobility shift assay [EMSA]) kit were purchased from Perkin
Elmer (Waltham, Massachusetts) and Promega (Madison,
Wisconsin), respectively.
|
431
Animals and treatments. Heterozygous knockout of COX-1 (COX1þ/) and homozygous knockout of COX-1 (COX-1/) mice were
generous gifts from Prof. Paul M. Vanhoutte (Department of
Pharmacology and Pharmacy, The University of Hong Kong)
(Tang et al., 2005). To ascertain the genotypes of the mice, DNA
collected from the mouse tail was subjected to PCR analysis as
previously described (Tang et al., 2005). The animals were
housed in a temperature-controlled room (21 6 1 C) with a 12-h
light/dark cycle (07:00 lights on, 19:00 lights off). Animals had
free access to chow (LabDiet 5010, St Louis, Missouri) and water.
The mice were divided into 6 groups (n ¼ 6 in each group): (1)
wild-type control mice with vehicle administration (olive oil or
phosphate buffer); (2) wild-type mice with CCl4 single-dose
treatment (50 ll/kg in olive oil; intraperitoneal injection); (3) heterozygous knockout of COX-1 (COX-1þ/) mice with CCl4 singledose treatment; (4) homozygous knockout of COX-1 (COX-1/)
mice with CCl4 single-dose treatment; (5) SC-560 pre-treated
wild-type mice with CCl4 single-dose treatment; and (6) PGE2
pre-treated COX-1/ mice CCl4 single-dose treatment. The dose
of single CCl4 treatment was 50 ll/kg (in olive oil, through i.p.
injection), based on our previous study (Chen et al., 2004).
Treatment duration of CCl4 was 8 h, which has been proven to
induce evidence of acute hepatic injury in previous studies
(Xiao et al., 2012a,b). In order to pharmacologically inhibit the
activity of COX-1 in mice, its specific inhibitor SC-560 was
injected 1 week before CCl4 intoxication (30 mg/kg in 0.1 M phosphate buffer containing 40% dimethyl sulfoxide (DMSO),
pH ¼ 7.4, 1 i.p. injection per day). On the last day, SC-560 was
given 30 min before CCl4 challenge (Choi et al., 2008). To test the
possible ameliorative effects of PGE2, synthetic PGE2 was intraperitoneally injected to COX-1/ mice at 3 mg/kg (6% ethanol in
saline) per day for 1 week before CCl4 challenge. Vehicle used
for PGE2 preparation showed no hepatic toxicity in mice (data
not shown). AA-861 (60 mg/kg dissolved in 0.1 ml DMSO) was
intraperitoneally injected 24 h before the D-galactosamine/lipopolysaccharide (Gal/LPS) administration (Li et al., 2014). After
8-h CCl4 intoxication, the mice were killed by an overdose of anesthesia according to the protocols approved by the Committee of
Animal Use for Research and Teaching at The University of Hong
Kong. The Laboratory Animal Unit of the University of Hong Kong is
fully accredited by the Association for Assessment and
Accreditation of Laboratory Animal Care International. Blood and
liver samples were collected for further analysis.
Serum and tissue sample processing and histological analysis. Serum
from whole blood sample was collected by centrifugation of
whole blood sample at 1000 g for 10 min at 4 C and stored
at 80 C. Liver tissue samples were fixed in 10% phosphatebuffered formalin, processed for histology, and embedded in
paraffin. Five-micrometer tissue sections were stained with hematoxylin and eosin for histological analysis under LEICA Qwin
Image Analyser (Leica Microsystems Ltd, Milton Keynes, UK).
Serum alanine aminotransferase and aspartate aminotransferase
assay. To evaluate the hepatic injury at the enzymatic level,
serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were measured by using ALT (SGPT)
reagent kit and AST (SGOT) reagent kit (Teco diagnostics,
Anaheim, California), respectively.
Terminal deoxynucleotidyl transferase-mediated dUTP-nick end labeling (TUNEL) assay. To demonstrate hepatic apoptosis, the TUNEL
assay, which detects 30 -hydroxyl ends in fragmented DNA as an
early event in the apoptotic cascade, was used. After dewaxing
432
|
TOXICOLOGICAL SCIENCES, 2015, Vol. 143, No. 2
and rehydration of the paraffin-embedded liver tissue sections,
sections were stained according to the manufacturer’s instructions regarding the TUNEL assay by using the in situ cell death
detection kit (Roche, Basel, Switzerland). Label solution without
terminal transferase was used in place of TUNEL reaction mixture for negative control. The positive immunostainings of
TUNEL were examined by light microscope (Zeiss Axiolab, Carl
Zeiss Inc, Germany) and analyzed by ImageJ software.
Measurement of malondialdehyde level. Levels of malondialdehyde
(MDA), the end-product of lipid peroxidation in all liver tissue
samples were determined by using a Bioxytech LPO-586 kit
(Oxis Research, Portland, Oregon). The reaction product was
measured spectrophotometrically at 586 nm. Standard curves
were constructed using 1,1,3,3-tetraethoxypropane as a standard. The MDA levels were normalized with corresponding
protein amounts determined by a Bio-Rad Protein Assay Kit
(Bio-Rad, Hercules, California) and expressed as percentage of
the control level.
RNA extraction and reverse-transcription quantitative polymerase
chain reaction. RNA extraction and reverse transcription were
performed as previously described according to the MIQE guideline (Bustin et al., 2009). The primer sequences and annealing
temperatures used in those Q-PCR reactions are listed in
Supplementary Table 1. Parallel amplification of glyceraldehyde-3-phosphate dehydrogenase was used as the internal control. Relative quantification was done by using the 2DDCt
method.
Western blot analysis. Western blot analyses of liver tissue
extracts were performed as described (Xiao et al., 2012a). The
ratio of the optical density of the protein product to the internal
control (b-actin) was obtained and was expressed as ratio or
percentage of the control value in the figures.
Enzyme-linked immunosorbent assay measurement. Protein expression of TNF-a, IL-1b, and monocyte chemotactic protein-1 (MCP1) were performed by using corresponding enzyme-linked
immunosorbent assay (ELISA) kits from PeproTech (PeproTech
Inc, Rocky Hill, New Jersey) according to user instructions. ELISA
assay of inducible nitric oxide synthase (iNOS) was done using
kits purchased from EIAab (Wuhan EIAab Science, Wuhan,
China). ELISA kit of 5-LO was bought from Ximei Bio (Shanghai,
China).
Determination of DNA-binding activity of NF-jB. The DNA binding
activity of NF-jB was performed by EMSA using the Gel Shift
Assay Systems from Promega as previously described (Xiao
et al., 2012a).
Measurements of serum cysteinyl leukotriene (LTC4, LTD4, and LTE4)
levels. To study the possible involvement of cysteinyl leukotriene after COX-1 knockout and CCl4 intoxication, serum levels
of LTC4, LTD4, and LTE4 have been measured by ELISA kit (Enzo
Life Sciences, Farmingdale, New York) and expressed as ratio or
percentage of the control value in the figures.
Statistical analysis. Data from each group were expressed as
means 6 SEM. Statistical comparison between groups was done
using the Kruskal-Wallis test followed by Dunn’s post hoc test
to detect differences in all groups. A P < 0.05 was considered to
be statistically significant (Prism 5.0, Graphpad software, Inc,
San Diego, California).
RESULTS
Genetic Deletion of COX-1 in Mice
To confirm the genotypes of COX-1 knockout mice in this study,
genotyping was conducted by PCR analysis of the genome DNA
extracted from the tail tissue of the mice. As shown in
Supplementary Figure 1, single band at 695 bp indicated homozygous deletion of COX-1 (COX-1/). While for the heterozygous group (COX-1þ/), two distinct bands at 628 bp and 695 bp
were obtained, which were from the original COX-1 allele and
an inserted sequence. In the wild-type group (COX-1þ/þ), one
band of 628 bp was obtained (Supplementary Fig. 1A). Western
blot targeting endogenous COX-1 expression further confirmed
the successful partial and complete genetic deletions of COX-1
(Supplementary Fig. 1B).
Deficiency of COX-1 Exacerbated Acute Liver Injury Induced by CCl4
After 8 h treatment with CCl4, the liver section of mice showed
evidence of necrosis and inflammation, particularly around the
centrilobular veins (Fig. 1A). Knockout of COX-1 exacerbated
the hepatic injury induced by CCl4, when compared with
CCl4-treated wild-type mice. Homozygous deletion of COX-1
(COX-1/) exhibited more severe hepatic injury than heterozygous deletion of COX-1 (COX-1þ/). Pharmacological inhibition
of COX-1 activity by SC-560 in wild-type mice also showed comparable hepatic injury to that of COX-1/ mice. However, when
COX-1/ mice were pre-treated with PGE2, the acute hepatic
injury induced by CCl4 was drastically attenuated (Fig. 1A).
Consistent with the histological results, CCl4 treatment induced
the elevation of both ALT and AST levels in mice serum, which
were further enhanced in the groups of COX-1þ/ and COX-1/
mice. SC-560 inhibition of COX-1 activity in wild-type mice
mimicked the effects of ALT and AST elevations as in the COX1/ group. Pre-treatment with PGE2 partially reversed the general hepatic injury as seen in the reduction of serum levels of
both ALT and AST (Figs. 1B and 1C).
COX-1 Is Essential for the Protection of Oxidative Stress Formation
As important endogenous antioxidant enzymes, the mRNA
expressions of catalase (CAT) and glutathione peroxidase (GPx)
were significantly reduced by CCl4 treatment. Knockout or
inhibition of activity of COX-1 in mice further reduced their
expressions, although there was no difference among COXþ/,
COX-1/, and SC-560 groups. Supplementation with PGE2 partially restored the expressions of CAT and GPx (Figs. 1D and 1E).
Consistent with the antioxidant enzyme results, the end-products of lipid peroxidation (MDA) and NO signaling (nitrotyrosine) showed significant increase after CCl4 treatment, with
further enhancement in their levels in the genetic deletion and
pharmacological inhibition COX-1 groups. PGE2 pre-treatment
partially or almost completely counter-acted the effects of COX1 deficiency as shown by the significant reduction of these endproducts (Figs. 1F and 1G).
Deletion of COX-1 Aggravated Pro-inflammatory Responses in the
Mice Liver
To examine the effects of COX-1 deletion or pharmacological
inhibitory activity on hepatic inflammation after CCl4 intoxication, expressions of typical markers for pro-inflammatory and
chemoattractive responses, including TNF-a, IL-1b, IL-6, COX-2,
iNOS, and MCP-1, were characterized by using ELISA. As
expected, CCl4 treatment induced expressional elevations of
these markers at the translational level. Such elevations were
further enhanced by the deletion or pharmacological inhibition
XIAO ET AL.
|
433
FIG. 1. Deficiency of COX-1 exacerbated hepatic injury and oxidative stress induced by CCl4 intoxication. A, Representative images of liver histology stained by H&E
method. Magnification, 200. B and C, Serum ALT and AST levels in each group of mice. D and E, Hepatic mRNA levels of endogenous antioxidant enzymes CAT and
GPx in each group of mice. F and G, Hepatic formation of oxidative stress productions nitrotyrosine and MDA in each group of mice (means 6 SEMs; n ¼ 6; different letters [eg, a vs b, b vs c] mean a P < 0.05 is statistically different between any two groups).
of COX-1. Pre-treatment with PGE2 partially abolished the deleterious effects of COX-1 deficiency (Figs. 2A–F). In addition, the
increased pro-inflammatory responses by CCl4 treatment were
partly attributed to increased number of Kupffer cells in the
liver, as shown by an increase expression level of ED1
(Supplementary Fig. 2). Pre-treatment with PGE2 significantly
reduced the hepatic ED1 level.
Apoptosis Was Worsened After COX-1 Deletion Through Both
Intrinsic and Extrinsic Apoptotic Pathways
Similar to oxidative stress and inflammation, apoptosis
is another common consequence of CCl4-induced acute
hepatic injury. TUNEL assay on liver sections from different
groups showed that the apoptotic ratio in wild-type control
mice liver was minimal. Administration of CCl4 induced hepatic
apoptosis, which was worsened in complete COX-1 knockout
and SC-560 pre-treatment mice, but not in partial COX-1 knockout group. PGE2 supplementation drastically attenuated the
apoptotic ratio to a very low level (Fig. 2G). To further study the
molecular mechanism of CCl4-induced apoptosis, we measured
key proteins regulating both intrinsic and extrinsic apoptotic
pathways. Eight-hour single treatment of CCl4 induced the protein expressions of intrinsic markers (cytochrome C, Bak1, Bcl-2,
and cleaved caspase-3) and extrinsic markers (FasL and cleaved
434
|
TOXICOLOGICAL SCIENCES, 2015, Vol. 143, No. 2
FIG. 2. Deletion of COX-1 exacerbated hepatic pro-inflammatory, chemoattractive, and apoptotic responses induced by CCl4 intoxication. Quantified ELISA data for protein expression of (A) TNF-a, (B) IL-1b, (C) IL-6, (D) COX-2, (E) iNOS, and (F) MCP-1 in each group of mice (means 6 SEMs; n ¼ 6; different letters [eg, a vs b, b vs c] mean a
P < 0.05 is statistically different between any two groups). (G) Representative images of hepatic apoptosis measured by TUNEL assay in the murine liver.
Magnification, 200.
caspase-8) (Figs. 3A and 3B). Partial deletion of COX-1 enhanced
most of these protein expressions, which were further elevated
by complete genetic deletion and pharmacological inhibition of
COX-1. In line with TUNEL results, supplementation with PGE2
significantly reduced the protein expressions of these apoptotic
markers (Figs. 3A and 3B).
NF-jB and C/EBP-a Were Involved in COX-1 Protection
Against CCl4 Toxicity
NF-jB and C/EBP-a are major regulators of inflammation and
apoptosis after CCl4 intoxication (Tao et al., 2012). In this study,
the DNA binding activity of NF-jB was drastically elevated
by CCl4 challenge. Knockout or pharmacological inhibition of
COX-1 showed higher elevation of NF-jB activity when compared with wild-type mice. PGE2 pre-treatment abolished such
effects (Fig. 3D). The elevation of NF-jB activity was accompanied with the protein degradation of its cytoplasmic inhibitor
IjBa (Fig. 3C). As a hepato-protective transcription factor against
CCl4-induced injury, Western blot results of C/EBP-a in nuclear
protein exhibited reverse expressional trends in the DNA binding activity of NF-jB (Fig. 3E).
Hepatic Protection and Liver Regeneration Were Impaired in COX-1
Knockout Mice After CCl4 Administration
Recently, the hepato-protective role of IL-6 family members—
oncostatin M (OSM) and its receptor OSMR were described
XIAO ET AL.
|
435
FIG. 3. Mechanisms for COX-1 deficiency-induced apoptosis and transcription factor activity. A, Representative images of Western blot assay results and corresponding
statistical data of markers for intrinsic apoptotic pathway cytochrome C, Bak1, Bcl-2, and cleaved caspase-3 in each group of mice (means 6 SEMs; n ¼ 3; different letters
[eg, a vs b, b vs c] mean a P < 0.05 is statistically different between any two groups). B, Representative images of Western blot assay results and corresponding statistical
data of markers for extrinsic apoptotic pathway FasL and cleaved caspase-8 in each group of mice (means 6 SEMs; n ¼ 3; different letters [eg, a vs b, b vs c] mean a
P < 0.05 is statistically different between any two groups). C and E, Representative images of Western blot assay results and corresponding statistical data of IjBa and
C/EBP-a in each group of mice (means 6 SEMs; n ¼ 4; different letters [eg, a vs b, b vs c] mean a P < 0.05 is statistically different between any two groups). D,
Representative images of radioisotope activity assay results and corresponding statistical data of NF-jB in each group of mice (means 6 SEMs; n ¼ 6; different letters
(eg, a vs b, b vs c) mean a P < 0.05 is statistically different between any two groups).
(Nakamura et al., 2004). In the current study, we found that CCl4
administration induced drastic elevation of both OSM and
OSMR expressions in the mice liver. Deletion of COX-1, however, significantly reduced the mRNA levels of OSM and OSMR,
indicating an impaired hepato-protective function COX-1 when
it was knockout (Figs. 4A and 4B). It is interesting that pre-treatment with PGE2 in COX-1/ mice did not influence the mRNA
level of OSM but restored the level of OSMR. In addition, as an
acute response to CCl4 challenge, liver regeneration-related signals (including TIMP1 and phosphorylated STAT3) were significantly increased. They were inhibited by the deletion or
pharmacological inhibition of COX-1 gene and partially restored
by the supplementation of PGE2 (Fig. 4C).
Exacerbation of CCl4-Induced Hepatic Injury Was Partially Through
the 5-Lipogenase (5-LO)/Cysteinyl Leukotrienes Pathway
The cysteinyl leukotrienes (LTC4, LTD4, and LTE4) are metabolites of arachidonic acid through the 5-LO pathway. LTC4 synthase is the enzyme limiting the formation of cysteinyl
leukotrienes (Yang et al., 2007). They are potent lipid mediators
in inflammation under several disease conditions, including
aspirin-intolerant asthma (Gyllfors et al., 2000), ischemia-reperfusion-injured liver (Yang et al., 2007), and CCl4-induced hepatic
fibrosis. To examine the involvement of the 5-LO/cysteinyl leukotrienes pathway in COX-1 knockout-exacerbated acute liver
injury model, the serum levels of LTC4/LTD4/LTE4 and the
mRNA expression of LTC4 synthase were measured. After CCl4
administration, the serum levels of LTC4/LTD4/LTE4 in wild-type
mice were significantly elevated when compared with vehicletreated wild-type mice. When COX-1 was knockout or pharmacologically inhibited, the serum levels of LTC4/LTD4/LTE4 were
further enhanced. Supplementation of PGE2 did not significantly reduce the cysteinyl leukotrienes levels when compared
with the CCl4-challenged COX-1/ mice (Fig. 4D). Consistently,
the mRNA expression level of LTC4 synthase and the protein
expression of 5-LO were peaked in COX-1 knockout mice when
compared with wild-type mice after CCl4 challenge. PGE2
administration did not significantly alter these changes (Figs. 4E
and 4F).
However, when 5-LO was specifically inhibited by its inhibitor AA-861 in COX-1/ mice, it significantly ameliorated the
severity of acute liver injury caused by the deficiency of COX-1.
When compared with CCl4 challenged COX-1/ mice, the pharmacological inhibition of 5-LO significantly improved hepatic
histology, reduced serum ALT level, hepatic MDA formation,
pro-inflammatory cytokine production, and restored OSMR
expression. In addition, AA-861 reduced the levels of cleaved
caspase-3 and -8, with restoration of phosphorylated STAT3
(Fig. 5).
DISCUSSION
The different expression patterns of COX-1 and COX-2 lead to
different physiological functions. COX-1 is considered as the
homeostatic stabilizer through continuous formation of PGs in
the liver whereas COX-2-derived PGs are mediators of
436
|
TOXICOLOGICAL SCIENCES, 2015, Vol. 143, No. 2
FIG. 4. Exacerbated acute liver injury by the knockout of COX-1 was partly attributed to a reduced expression level of hepato-protective molecules, liver regeneration
proteins, and enhanced 5-LO/cysteinyl leukotrienes pathway. Quantified data for the mRNA expression of (A) OSM and (B) OSMR in each group of mice (means 6 SEM;
n ¼ 6; different letters [eg, a vs b, b vs c] mean a P < 0.05 is statistically different between any two groups). (C) Representative images of Western blot assay results and
corresponding statistical data of TIMP1, phosphorylated STAT3, and total STAT3 (means 6 SEMs; n ¼ 4; different letters mean a P < 0.05 is statistically different between
any two groups). (D–F) Quantified data for contents of serum LTC4/LTD4/LTE4, LTC4S mRNA level, and 5-LO protein level in each group of mice (means 6 SEM; n ¼ 6; different letters [eg, a vs b, b vs c] mean a P < 0.05 is statistically different between any two groups).
XIAO ET AL.
|
437
FIG. 5. Inhibition of 5-LO ameliorated CCl4-induced acute liver injury. A, Representative images of liver histology stained by H&E method. Magnification, 200. B,
Serum ALT level; C, hepatic MDA content; D, hepatic TNF-a protein level; and E, hepatic OSMR mRNA level in each group of mice. F, Representative images of Western
blot assay results and corresponding statistical data of cleaved caspase-3, cleaved caspase-8, phosphorylated STAT3, and total STAT3 (means 6 SEMs; n ¼ 3; different
letters mean a P < 0.05 is statistically different between any two groups).
pathological conditions such as inflammation and carcinogenesis (Wu et al., 2009). Currently, little is known about the role of
COX-1 in inflammatory liver injury response. In the present
study, we found that the deficiency or inhibition of enzymatic
activity of COX-1 exacerbated the severity of CCl4-induced acute
liver injury, including elevated serum aminotransferases levels,
increased necrosis, and apoptosis in the liver, enhanced hepatic
oxidative stress, and pro-inflammatory responses. Homozygous
COX-1 knockout mice showed more severe acute liver injury
than heterozygous deletion, which further proved the direct
link of COX-1 and liver injury. Hence, we proposed 3 possible
mechanisms that lead to increased severity of hepatic damage
by COX-1 deficiency: (1) COX-1 catalyzed PGE2 was the key protector against CCl4-induced damage, since pre-treatment with
synthetic PGE2 in homozygous COX-1 knockout mice significantly ameliorated acute liver injury; (2) deficiency of COX-1
reduced the expression level of hepato-protective molecules
(eg, OSM and OSMR) and liver regeneration markers (TIMP1 and
STAT3) after CCl4 challenge; and (3) metabolites (cysteinyl leukotrienes) of arachidonic acid by 5-LO pathway, which were
important mediators of hepatic inflammatory responses, were
further enhanced by the deficiency or pharmacological inhibition of COX-1.
The anti-inflammatory cascade related to the COX-PGE2-EP
receptor signaling has been documented in several reports. In
Helicobacter pylori-induced gastritis mice, PGE2 derived from
COX-1 was involved in gastric mucosal inflammation through
the inhibition of TNF-a (Tanigawa et al., 2004). Another study
demonstrated that COX-1 deficiency was detrimental to the
host defense against influenza infection. The mechanism was
closely associated with reduced production of PGE2 and leukotrienes (LTC4, LTD4, and LTE4) following infection, which was
consistent with our current findings in the liver (Carey et al.,
2005). Activation of PG E receptor 4 (EP4) was also documented
to play an anti-inflammatory property in cardiovascular and
other inflammatory diseases, making EP4 an attractive target
for the attenuation of syndromes with inflammation (Tang
et al., 2012). Furthermore, the production of PGEs was required
during liver regeneration in response to liver injury through the
activation of cAMP response element-binding protein (CREB)
(Rudnick et al., 2001). In the current study, the number of activated Kupffer cells and chemokine expression were reduced by
438
|
TOXICOLOGICAL SCIENCES, 2015, Vol. 143, No. 2
FIG. 6. Diagrammatic illustration of the protective mechanisms of COX-1 against CCl4-induced acute liver injury.
PGE2 administration (shown by ED1 staining) (Schemmer et al.,
1998), the vital role of PGE2 from COX-1 production has been
demonstrated. Furthermore, our previous study demonstrated
that in acute liver injury model, there was an inverse expression
correlation between COX-1 and COX-2, which was consistent
with the current findings that deficiency of COX-1 induced
higher level of hepatic inflammatory responses, since COX-2 is
one of the key inflammatory markers (Liong et al., 2012).
OSM and its cognate receptor are members of the IL-6 family,
which have been shown to be hepato-protective in many systems, including the liver, through the activation of STAT3
(Pritchard et al., 2010). For instance, OSMR knockout mice
showed persistent necrosis after the intoxication of CCl4, with
reduced activation level of STAT3 (Nakamura et al., 2004). In
addition, OSM gene therapy ameliorated hepatic injury induced
by dimethylnitrosamine in rat (Hamada et al., 2007). Recently, it
was found that OSM could be directly produced in Kupffer cells
by PGE2 in the liver, which further explained the essential protective function of COX-1 and its production of PGE2 and downstream OSM were partly derived from the local response of
Kupffer cells. Indeed, an unexplained result that supplementation with PGE2 recovered the OSMR expression but not OSM in
the liver warrants further investigation.
Arachidonic acid can be metabolized to produce PGs (including PGE2) either through COX- or 5-LO-mediated pathway. When
COX-1 expression is decreased, the expression of 5-LO probably
increases as a compensatory mechanism for the maintenance of
PGs production (Holgate et al., 2003). Since the metabolites of
arachidonic acid, such as cysteinyl leukotrienes, are important
mediators of inflammatory responses (McGeehan and Bush,
2002), deficiency of COX-1 may provoke hepatic oxidative stress
and inflammation partly through the enhanced 5-LO/cysteinyl
leukotrienes pathway. Very importantly, after CCl4 challenge,
when 5-LO was inhibited by AA-861, the hepatic injury, levels of
ALT, oxidative stress, inflammation, and apoptosis were
significantly ameliorated in COX1/ mice when compared with
the same mice without 5-LO inhibition (Fig. 5). This could partly
explain why pre-treatment with PGE2 failed to reduce the elevated levels of the 5-LO/cysteinyl leukotrienes pathway (Fig. 4). A
recent study showed similar results that inhibition of 5-LO pathway by AA-861 attenuated acute liver failure in rats by inhibiting
macrophage activation (Li et al., 2014).
Taken together, deficiency or pharmacological inhibition of
COX-1 in mice aggravated CCl4-induced acute liver injury.
Reduced production of PGE2, decreased expression of OSM/
OSMR and liver regeneration markers, and enhanced 5-LO/cysteinyl leukotrienes pathway were mainly responsible for the
detrimental effects in the liver of COX-1-deficient mice. Pretreatment with PGE2 or 5-LO inhibitor ameliorated the exacerbated liver injury caused by COX-1 deficiency by down-regulating the markers of oxidative stress, inflammation, necrosis, and
apoptosis through both intrinsic and extrinsic pathways (Fig. 6).
We proposed that a basal expression COX-1 is critical for the
protection of the liver against acute injury.
SUPPLEMENTARY DATA
Supplementary data are available online at http://toxsci.
oxfordjournals.org/.
FUNDING
Small Project Funding (in part), University Research
Committee, The University of Hong Kong and General
Research Fund, University Grant Council, Hong Kong SAR.
ACKNOWLEDGMENTS
The authors declare that they have no competing interests.
XIAO ET AL.
REFERENCES
Bustin, S. A., Benes, V., Garson, J. A., Hellemans, J., Huggett, J.,
Kubista, M., Mueller, R., Nolan, T., Pfaffl, M. W., Shipley, G. L.,
et al. (2009). The MIQE guidelines: minimum information for
publication of quantitative real-time PCR experiments. Clin.
Chem. 55, 611–622.
Carey, M. A., Bradbury, J. A., Seubert, J. M., Langenbach, R.,
Zeldin, D. C., and Germolec, D. R. (2005). Contrasting effects
of cyclooxygenase-1 (COX-1) and COX-2 deficiency on the
host response to influenza A viral infection. J. Immunol. 175,
6878–6884.
Cha, Y. I., and DuBois, R. N. (2007). NSAIDs and cancer prevention: targets downstream of COX-2. Ann. Rev. Med. 58,
239–252.
Chandrasekharan, N. V., Dai, H., Roos, K. L., Evanson, N. K.,
Tomsik, J., Elton, T. S., and Simmons, D. L. (2002). COX-3, a
cyclooxygenase-1 variant inhibited by acetaminophen and
other analgesic/antipyretic drugs: cloning, structure, and expression. Proc. Natl Acad. Sci. U. S. A. 99, 13926–13931.
Chen, J. H., Tipoe, G. L., Liong, E. C., So, H. S., Leung, K. M., Tom,
W. M., Fung, P. C., and Nanji, A. A. (2004). Green tea polyphenols prevent toxin-induced hepatotoxicity in mice by downregulating inducible nitric oxide-derived prooxidants. Am. J.
Clin. Nutr. 80, 742–751.
Choi, S. H., Langenbach, R., and Bosetti, F. (2008). Genetic deletion or pharmacological inhibition of cyclooxygenase-1 attenuate lipopolysaccharide-induced inflammatory response
and brain injury. FASEB J. 22, 1491–1501.
Daikoku, T., Tranguch, S., Trofimova, I. N., Dinulescu, D. M.,
Jacks, T., Nikitin, A. Y., Connolly, D. C., and Dey, S. K. (2006).
Cyclooxygenase-1 is overexpressed in multiple genetically
engineered mouse models of epithelial ovarian cancer.
Cancer Res. 66, 2527–2531.
Gill, R. Q., and Sterling, R. K. (2001). Acute liver failure. J. Clin.
Gastroenterol. 33, 191–198.
Gyllfors, P., Bochenek, G., Overholt, J., Drupka, D., Kumlin, M.,
Sheller, J., Nizankowska, E., Isakson, P. C., Mejza, F.,
Lefkowith, J. B., et al. (2000). Hepatocyte-derived cysteinyl
leukotrienes modulate vascular tone in experimental cirrhosis. Gastroenterology 119, 794–805.
Hamada, T., Sato, A., Hirano, T., Yamamoto, T., Son, G., Onodera,
M., Torii, I., Nishigami, T., Tanaka, M., Miyajima, A., et al.
(2007). Oncostatin M gene therapy attenuates liver damage
induced by dimethylnitrosamine in rats. Am. J. Pathol. 171,
872–881.
Harris, S. G., Padilla, J., Koumas, L., Ray, D., and Phipps, R. P.
(2002). Prostaglandins as modulators of immunity. Trends
Immunol. 23, 144–150.
Holgate, S. T., Peters-Golden, M., Panettieri, R. A., and
Henderson, W. R., Jr. (2003). Roles of cysteinyl leukotrienes in
airway inflammation, smooth muscle function, and remodeling. J. Allergy Clin. Immunol. 111, S18–S34; discussion
S34–S36.
Khan, R. A., Khan, M. R., and Sahreen, S. (2012). CCl4-induced
hepatotoxicity: protective effect of rutin on p53, CYP2E1 and
the antioxidative status in rat. BMC Complement. Altern. Med.
12, 178.
Lampiasi, N., Fodera, D., D’Alessandro, N., Cusimano, A.,
Azzolina, A., Tripodo, C., Florena, A. M., Minervini, M. I.,
Notarbartolo, M., Montalto, G., et al. (2006). The selective cyclooxygenase-1 inhibitor SC-560 suppresses cell proliferation
and induces apoptosis in human hepatocellular carcinoma
cells. Int. J. Mol. Med. 17, 245–252.
|
439
Li, L., Liu, Y. R., Gao, S., Li, J. F., Li, S. S., Zhang, D. D., Liu, S., Bai, L.,
Zheng, S. J., Duan, Z. P., et al. (2014). Inhibition of
5-lipoxygenase pathway attenuates acute liver failure by
inhibiting macrophage activation. J. Immunol. Res. 2014,
697560.
Liong, E. C., Xiao, J., Lau, T. Y., Nanji, A. A., and Tipoe, G. L. (2012).
Cyclooxygenase inhibitors protect D-galactosamine/lipopolysaccharide induced acute hepatic injury in experimental
mice model. Food Chem. Toxicol. 50, 861–866.
Liu, C., Tao, Q., Sun, M., Wu, J. Z., Yang, W., Jian, P., Peng, J., Hu,
Y., Liu, C., and Liu, P. (2010). Kupffer cells are associated with
apoptosis, inflammation and fibrotic effects in hepatic fibrosis in rats. Lab. Invest. 90, 1805–1816.
Martin-Sanz, P., Mayoral, R., Casado, M., and Bosca, L. (2010).
COX-2 in liver, from regeneration to hepatocarcinogenesis:
what we have learned from animal models? World J.
Gastroenterol. 16, 1430–1435.
McGeehan, M., and Bush, R. K. (2002). The mechanisms of aspirin-intolerant asthma and its management. Curr. Allergy
Asthma Rep. 2, 117–125.
Nakamura, K., Nonaka, H., Saito, H., Tanaka, M., and Miyajima,
A. (2004). Hepatocyte proliferation and tissue remodeling is
impaired after liver injury in oncostatin M receptor knockout
mice. Hepatology 39, 635–644.
Pritchard, M. T., Cohen, J. I., Roychowdhury, S., Pratt, B. T., and
Nagy, L. E. (2010). Early growth response-1 attenuates liver injury and promotes hepatoprotection after
carbon tetrachloride exposure in mice. J. Hepatol. 53,
655–662.
Rudnick, D. A., Perlmutter, D. H., and Muglia, L. J. (2001).
Prostaglandins are required for CREB activation and cellular
proliferation during liver regeneration. Proc. Natl Acad. Sci.
U. S. A. 98, 8885–8890.
Rutherford, A., and Chung, R. T. (2008). Acute liver failure: mechanisms of hepatocyte injury and regeneration. Semin. Liver
Dis. 28, 167–174.
Schemmer, P., Schoonhoven, R., Swenberg, J. A., Bunzendahl, H.,
and Thurman, R. G. (1998). Gentle in situ liver manipulation
during organ harvest decreases survival after rat liver transplantation: role of Kupffer cells. Transplantation 65,
1015–1020.
Spencer, S. L., and Sorger, P. K. (2011). Measuring and modeling
apoptosis in single cells. Cell 144, 926–939.
Tang, E. H., Ku, D. D., Tipoe, G. L., Feletou, M., Man, R. Y., and
Vanhoutte, P. M. (2005). Endothelium-dependent contractions occur in the aorta of wild-type and COX2/ knockout
but not COX1/ knockout mice. J. Cardiovasc. Pharmacol. 46,
761–765.
Tang, E. H., Libby, P., Vanhoutte, P. M., and Xu, A. (2012). Antiinflammation therapy by activation of prostaglandin EP4 receptor in cardiovascular and other inflammatory diseases.
J. Cardiovasc. Pharmacol. 59, 116–123.
Tanigawa, T., Watanabe, T., Hamaguchi, M., Sasaki, E.,
Tominaga, K., Fujiwara, Y., Oshitani, N., Matsumoto, T.,
Higuchi, K., and Arakawa, T. (2004). Anti-inflammatory effect
of two isoforms of COX in H. pylori-induced gastritis in mice:
possible involvement of PGE2. Am. J. Physiol. Gastrointest. Liver
Physiol. 286, G148–G156.
Tao, L. L., Cheng, Y. Y., Ding, D., Mei, S., Xu, J. W., Yu, J., Ou-Yang,
Q., Deng, L., Chen, Q., Li, Q. Q., et al. (2012). C/EBP-alpha
ameliorates CCl(4)-induced liver fibrosis in mice through
promoting apoptosis of hepatic stellate cells with little apoptotic effect on hepatocytes in vitro and in vivo. Apoptosis 17,
492–502.
440
|
TOXICOLOGICAL SCIENCES, 2015, Vol. 143, No. 2
Wang, D., and Dubois, R. N. (2006). Prostaglandins and cancer.
Gut 55, 115–122.
Wu, B., Zeng, L., Lin, Y., Wen, Z., Chen, G., Iwakiri, R., and
Fujimoto, K. (2009). Downregulation of cyclooxygenase-1 is
involved in gastric mucosal apoptosis via death signaling in
portal hypertensive rats. Cell Res. 19, 1269–1278.
Xiao, J., Liong, E. C., Ching, Y. P., Chang, R. C., So, K. F., Fung, M. L.,
and Tipoe, G. L. (2012a). Lycium barbarum polysaccharides
protect mice liver from carbon tetrachloride-induced oxidative stress and necroinflammation. J. Ethnopharmacol. 139,
462–470.
Xiao, J., Liong, E. C., Ling, M. T., Ching, Y. P., Fung, M. L., and
Tipoe, G. L. (2012b). S-allylmercaptocysteine reduces carbon
tetrachloride-induced hepatic oxidative stress and necroinflammation via nuclear factor kappa B-dependent pathways
in mice. Eur. J. Nutr. 51, 323–333.
Yang, S. L., Huang, X., Chen, H. F., Xu, D., Chen, L. J., Kong, Y., and
Lou, Y. J. (2007). Increased leukotriene c4 synthesis accompanied enhanced leukotriene c4 synthase expression and activities of ischemia-reperfusion-injured liver in rats. J. Surg. Res.
140, 36–44.