Date post: | 05-Mar-2023 |
Category: |
Documents |
Upload: | michiganstate |
View: | 0 times |
Download: | 0 times |
TOXICOLOGICAL SCIENCES 100(1), 267–280 (2007)
doi:10.1093/toxsci/kfm209
Advance Access publication August 13, 2007
The Role of Tumor Necrosis Factor Alpha in Lipopolysaccharide/Ranitidine-Induced Inflammatory Liver Injury
Francis F. Tukov,*,† James P. Luyendyk,*,† Patricia E. Ganey,*,† and Robert A. Roth*,†,1
*Center for Integrative Toxicology and †Department of Pharmacology and Toxicology, Michigan State University, East Lansing, Michigan 48824
Received March 20, 2007; accepted July 26, 2007
Exposure to a nontoxic dose of bacterial lipopolysaccharide
(LPS) increases the hepatotoxicity of the histamine-2 (H2) re-
ceptor antagonist, ranitidine (RAN). Because some of the
pathophysiologic effects associated with LPS are mediated
through the expression and release of inflammatory mediators
such as tumor necrosis factor alpha (TNF), this study was
designed to gain insights into the role of TNF in LPS/RAN
hepatotoxicity. To determine whether RAN affects LPS-induced
TNF release at a time near the onset of liver injury, male
Sprague-Dawley rats were treated with 2.5 3 106 endotoxin units
(EU)/kg LPS or its saline vehicle (iv) and 2 h later with either 30
mg/kg RAN or sterile phosphate-buffered saline vehicle (iv). LPS
administration caused an increase in circulating TNF concentra-
tion. RAN cotreatment enhanced the LPS-induced TNF increase
before the onset of hepatocellular injury, an effect that was not
produced by famotidine, a H2-receptor antagonist without
idiosyncrasy liability. Similar effects were observed for serum
interleukin (IL)-1beta, IL-6, and IL-10. To determine if TNF plays
a causal role in LPS/RAN-induced hepatotoxicity, rats were given
either pentoxifylline (PTX; 100 mg/kg, iv) to inhibit the synthesis
of TNF or etanercept (Etan; 8 mg/kg, sc) to impede the ability of
TNF to reach cellular receptors, and then they were treated with
LPS and RAN. Hepatocellular injury, the release of inflammatory
mediators, hepatic neutrophil (PMN) accumulation, and bio-
markers of coagulation and fibrinolysis were assessed. Pretreat-
ment with either PTX or Etan resulted in the attenuation of liver
injury and diminished circulating concentrations of TNF, IL-1b,IL-6, macrophage inflammatory protein-2, and coagulation/
fibrinolysis biomarkers in LPS/RAN-cotreated animals. Neither
PTX nor Etan pretreatments altered hepatic PMN accumulation.
These results suggest that TNF contributes to LPS/RAN-induced
liver injury by enhancing inflammatory cytokine production and
hemostasis.
Key Words: tumor necrosis factor alpha; inflammation; liver
injury; lipopolysaccharide; ranitidine; adverse drug reactions;
coagulation; hemostasis; hepatotoxicity.
Previous studies in rats showed that modest inflammation
triggered by bacterial lipopolysaccharide (LPS) decreases the
threshold for xenobiotic hepatotoxicity (reviewed in Roth
et al., 1997). One of the xenobiotics for which this is true is the
histamine-2 (H2) receptor antagonist, ranitidine (RAN; Luyendyk
et al., 2003). RAN has been associated with idiosyncratic
adverse drug reactions (IADRs) in people, with the liver as
a frequent target. Idiosyncratic hepatotoxicity occurs in less
than 0.1% of people taking RAN (Vial et al., 1991), and most
RAN-induced liver reactions are mild and reversible (Ribeiro
et al., 2000). In contrast, another H2-receptor antagonist,
famotidine (FAM), is not associated with idiosyncratic hepa-
totoxicity in humans: very few reports of FAM-associated
hepatotoxicity have been published (Ament et al., 1994;
Hashimoto et al., 1994; Jimenez-Saenz et al., 2000), and the
contribution of FAM to liver injury in these cases has been
questioned (Luyendyk et al., 2003).
Although it did not produce liver injury in healthy
experimental animals, RAN (30 mg/kg) was rendered hepato-
toxic in rats undergoing a mild inflammatory response induced
by LPS (Luyendyk et al., 2003). By contrast, a dose of FAM
that was equal in pharmacological efficacy (6 mg/kg) did not
interact with LPS to produce liver injury. LPS/RAN-cotreated
animals developed midzonal hepatocellular necrosis and a
liver-related clinical chemistry pattern resembling human cases
of RAN idiosyncrasy. Hepatocellular oncotic necrosis in LPS/
RAN-treated rats was preceded by hepatic neutrophil (PMN)
accumulation, and depletion of circulating PMNs attenuated
LPS/RAN-induced liver injury, suggesting a critical role for
these cells in this response (Luyendyk et al., 2005). In other
studies, activation of the coagulation system was shown to be
crucial to the pathogenesis of LPS/RAN hepatotoxicity
(Luyendyk et al., 2004b).
These results support the hypothesis that inflammation and
coagulation initiated by exposure to small amounts of LPS are
involved in the mechanism of liver injury from LPS/RAN
cotreatment. Evidence from both in vivo and in vitro studies
suggests that tumor necrosis factor alpha (TNF) is a critical and
proximal mediator of the inflammatory and hemostatic path-
ways stimulated by LPS (Aderka et al., 1992; Beutler and
Cerami, 1988; van der Poll et al., 1997). Furthermore, at a
1 To whom correspondence should be addressed at Department of
Pharmacology and Toxicology, National Food Safety and Toxicology Center,
221 Food Safety and Toxicology Building, Michigan State University, East
Lansing, MI 48824. Fax: (517) 432-2310. E-mail: [email protected].
� The Author 2007. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved.For Permissions, please email: [email protected]
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
relatively larger dose of LPS, liver injury is caused by complex
mechanisms involving the interaction of numerous soluble
mediators such as TNF and inflammatory cells, including
Kupffer cells (KCs), which are thought to be the major source
of TNF in liver (Ganey and Roth, 2001; Hewett et al., 1992,
1993). These findings raise the possibility that TNF contributes
to liver damage caused by cotreatment with LPS and RAN.
The purpose of this study was to characterize the expression
of inflammatory mediators during LPS/RAN interaction in vivoand in vitro and to determine the role of TNF in the ability of
LPS to render a nontoxic dose of RAN hepatotoxic. Toward
these ends, the effects of RAN and LPS on production of TNF
and other inflammatory factors in rats and in KC/hepatocyte
(KC/HPC) cocultures were determined and compared with
those elicited by FAM. Finally, by neutralizing TNF, the
hypothesis that TNF is a crucial mediator in LPS/RAN-induced
hepatotoxicity was tested.
MATERIALS AND METHODS
Materials. LPS derived from Escherichia coli serotype O55:B5 (Sigma
Chemical Co., St Louis, MO) with an activity of 13 3 106 endotoxin units
(EU)/mg was used for these studies. This activity was determined using
a limulus amebocyte lysate colorimetric end-point assay (kit 50-6480)
purchased from Cambrex Corp. (East Rutherford, NJ). Pentoxifylline (PTX),
and fetal bovine serum (FBS) were obtained from Sigma Chemical Co.
Etanercept (Etan; Enbrel; Immunex Corporation, Thousand Oaks, CA) was
obtained from the Michigan State University Clinical Center Pharmacy.
Antibiotic/antimycotic solution, gentamicin and RPMI 1640 (supplemented
with 2mM L-glutamine) were purchased from Gibco BRL (Rockville, MD).
The WEHI-13VAR mouse fibrosarcoma cell line was obtained from ATCC
(Manassas, VA). Unless otherwise stated, all other materials were purchased
from Sigma Chemical Co.
Animals. Male Sprague-Dawley rats (Crl:CD [SD]IGS BR; Charles River,
Portage, MI) weighing 250–350 g were used for these studies. Animals were
fed standard chow (Rodent Chow/Tek 8640; Harlan Teklad, Madison, WI) and
allowed access to water ad libitum. They were allowed to acclimate for 1 week
in a 12-h light/dark cycle before use. All procedures on animals followed the
guidelines for humane treatment set by the American Association of Laboratory
Animal Sciences and the University Laboratory Animal Research Unit at
Michigan State University.
Assessment of the effects of RAN on LPS-induced TNF release
in vivo. Rats fasted for 24 h were given 2.5 3 106 EU/kg LPS or its saline
vehicle, iv via a tail vein. Two hours later, 30 mg RAN/kg or sterile phosphate-
buffered saline (PBS) vehicle was administered through a tail vein. At 0, 2, 3, 6,
or 12 h after RAN administration, rats were anesthetized with sodium
pentobarbital (50 mg/kg, ip) and killed by exsanguination. For the collection of
plasma, blood was collected into vacutainer tubes (Becton Dickenson, Franklin
Lakes, NJ, USA) containing sodium citrate (final concentration, 3.8%).
Remaining blood was allowed to clot at room temperature, and serum was
collected and stored at � 80�C until use. TNF concentration was determined by
enzyme-linked immunosorbent assay (ELISA; see below). In a separate study
to examine inflammatory mediator release before the onset of injury, rats were
treated as above except that a group given 6 mg/kg FAM instead of RAN was
included, and the animals were killed 1, 2, or 3 h later. For these studies, RAN
and FAM were administered at doses that have equal pharmacological efficacy
in humans (Lin, 1991; Scarpignato et al., 1987) to simulate the ratio of doses
taken by people. As a result, the dose of RAN was fivefold greater than the dose
of FAM. We have shown previously that cotreatment of rats with LPS and
a dose of FAM equimolar or equally efficacious to the dose of RAN used in
these studies does not result in liver injury (Luyendyk et al., 2003).
Determination of PTX efficacy. PTX, a xanthine oxidase inhibitor, has
been used as a tool for exploring the role of TNF in xenobiotic-induced
hepatotoxicity (Sneed et al., 2000; Yee et al., 2003). PTX inhibits transcription
of TNF mRNA through a mechanism that involves phosphodiesterase
inhibition and the elevation of intracellular cyclic adenosine monophosphate
(Doherty et al., 1991; Schandene et al., 1992). Rats fasted for 24 h received
PTX (100 mg/kg) or its saline vehicle intravenously 1 h before treatment with
LPS (2.5 3 106 EU/kg) or its saline vehicle, iv. This treatment protocol for
PTX has been shown to prevent the LPS-induced rise in plasma TNF
concentration (Hewett et al., 1993). Two hours after LPS or vehicle exposure,
rats were anesthetized with sodium pentobarbital and killed by exsanguination.
Blood was allowed to clot at room temperature, and serum was collected and
stored at � 80�C until use. TNF concentration was determined using ELISA.
Determination of Etan efficacy. Rats fasted for 24 h received 8 mg Etan/kg
or its sterile water vehicle sc 1 h before LPS (2.5 3 106 EU/kg) or its saline
vehicle. Two hours later, they were anesthetized and bled as described above.
This Etan treatment has been shown to inactivate TNF activity (Geier et al.,
2003). A cytolytic cell assay (WEHI assay) was used to measure serum TNF
activity as described previously (Eskandari et al., 1990; Espevik and Nissen-
Meyer, 1986; Hewett et al., 1993; Sneed et al., 2000) with slight modifications.
Briefly, dilutions of serum samples from rats previously exposed to either veh/
LPS or Etan/LPS were assayed in triplicate for cytotoxic activity in WEHI-
13VAR mouse fibrosarcoma cells seeded at a density of 5 3 104 cells/well in
RPMI 1640 medium supplemented with 10% FBS, 1% antibiotic/antimycotic
solution, and 50 lg/ml gentamicin. The extent of cell death was measured with
3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide using a Bio-Tek
plate reader. Absorbance was read at 570 nm. A recombinant rat TNF (BD
Biosciences, San Diego, CA) was used as standard to calculate TNF activity in
the samples.
Assessment of LPS/RAN-induced hepatotoxicity in rats after PTX or Etan
pretreatment. Rats were fasted for 24 h before experiments. One hour before
administration of LPS (2.5 3 106 EU/kg) or its saline vehicle, rats received
either PTX (100 mg/kg) or its saline vehicle intravenously or Etan (8 mg/kg) or
its water vehicle sc. Two hours after LPS treatment, they were treated with
either 30 mg RAN/kg or its vehicle, and 6 h later they were anesthetized, bled,
and killed as described above. Blood was allowed to clot at room temperature,
and serum was collected by centrifugation and assayed for alanine
aminotransferase (ALT) activity using a diagnostic kit (ALT [GPT]; Ref:
TR7 111-125, Infinity; Thermo Electron, Melbourne, Australia).
Histopathology. Livers were fixed by immersion in 10% neutral-buffered
formalin for at least 3 days before being processed for histologic analysis.
Formalin-fixed liver samples from the left lateral liver lobe (3 samples/rat) were
embedded in paraffin, sectioned at 5 lm, stained for hematoxylin and eosin,
and examined by light microscopy for lesion size as described previously by
Luyendyk et al., (2003). All tissues were examined without knowledge of
treatment. The nature of lesions did not differ qualitatively among groups, and
each was assigned a score of 0 (no injury) to 6 based on increasing size of the
lesions examined at 1003 magnification.
Concentrations of cytokines, plasminogen activator inhibitor-1, and
thrombin-antithrombin dimers. Serum TNF concentration was measured
using a commercial kit (Rat TNF ELISA Kit II, BD OptEIA; BD Biosciences)
with recombinant rat TNF as standard. Measurements were performed in
duplicate. The concentrations of interleukin (IL)-1b, IL-6, IL-10, interferon-
gamma, and macrophage inflammatory protein-2 (MIP-2) in serum were
determined using enzyme immunoassay (EIA) kits obtained from BioSource
International, Inc. (Camarillo, CA). The assays were performed according to
the manufacturer’s instructions, and all samples were assayed in duplicate. The
concentration of plasminogen activator inhibitor-1 (PAI-1) in serum was
evaluated using an ELISA purchased from American Diagnostica, Inc.
268 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
(Greenwich, CT). This ELISA measures total PAI-1 including active, inactive,
and tissue plasminogen activator–complexed forms. Plasma thrombin-anti-
thrombin (TAT) concentration was determined using an EIA kit (Enzygnost
TAT micro; #OWMG15) from Dade Behring Inc. (Deerfield, IL). Serum
concentrations of prostaglandin E2 (PGE2) were determined using a commer-
cially available EIA kit (PGE2 Express EIA Kit; Cayman Chemical, Ann Arbor,
MI). EIA was performed according to the manufacturer’s instructions, and all
samples were assayed in duplicate.
Detection of hepatic fibrin. Fibrin immunohistochemistry and quantifica-
tion were performed as described previously (Copple et al., 2002; Luyendyk
et al., 2004a). Eight micrometer thick sections of frozen liver were fixed in 10%
buffered formalin containing 2% acetic acid for 30 min at room temperature.
This fixation protocol solubilizes all fibrinogen and fibrin species except for
cross-linked fibrin. Sections were blocked with PBS containing 10% horse
serum (i.e., blocking solution; Vector Laboratories, Burlingame, CA) for
30 min, and this was followed by incubation overnight at 4�C with goat antirat
fibrinogen diluted (1:1000; ICN Pharmaceuticals, Aurora, OH) in blocking
solution. Next, sections were incubated for 3 h with donkey antigoat secondary
antibody conjugated to Alexa 594 (1:1000; Molecular Probes, Eugene, OR) in
blocking solution for 3 h. Sections were washed three times, 5 min each, with
PBS and visualized using an Olympus AX-80T fluorescence microscope
(Olympus, Lake Success, NY). Ten randomly chosen digital images (1003
magnification) were captured using a SPOT II camera and SPOT advanced
software (Diagnostic Instruments, Sterling Heights, MI). Samples were coded
such that the evaluator was not aware of treatment. Each digital image
encompassed a total area of 1.4 mm2 and contained several centrilobular and
periportal regions. Quantification of immunostaining was performed with Scion
Image Beta 4.0.2 (Scion Corporation, Frederick, MD) using the method
described by Copple et al. (2002). Results from 10 random fields analyzed per
liver section were averaged and counted as a replicate, that is, each replicate
represents a different rat.
Evaluation of hepatic PMN accumulation. PMN immunohistochemistry
was performed on formalin-fixed liver sections as described previously
(Luyendyk et al., 2004a; Yee et al., 2003). Briefly, the paraffin-embedded
liver tissue was cut into 6 lm thick slices (three serial liver sections per slide).
Paraffin was removed from the liver tissues with xylene before staining. PMNs
within liver sections were stained with a rabbit anti-PMN Ig isolated from
serum of rabbits immunized with rat PMNs as described by Hewett et al.
(1992). After incubation with the primary antibody, tissue sections were
incubated with biotinylated goat antirabbit IgG, avidin-conjugated alkaline
phosphatase, and Vector Red substrate to stain PMNs. The number of PMNs in
10–20 randomly selected, high-power fields (4003) was counted for each liver.
The average of these numbers was calculated, and hepatic PMN accumulation
is presented as the mean for each treatment group.
Isolation and culture of rat HPCs and KCs. HPCs and KCs were isolated
from the same rat liver as described by Tukov et al. (2006). After the isolation
of HPCs, cell viability was assessed by trypan blue exclusion. If cells were
greater than 80% viable, they were kept on ice for later use, at which time the
HPCs were plated in 24-well culture plates at a density of 2.5 3 105 cells/well.
Similarly, the viability of the isolated KCs was determined by trypan blue
exclusion and was usually > 90%. The cell concentration was adjusted to 1 3
106 viable cells/ml, and the KCs (1 3 106 cells/well) were plated in 24-well
plastic culture plates (Costar, Cambridge, MA) for 20 min at 37�C in
a humidified incubator (95% air/5% CO2). After this time, nonadherent cells
were removed by replacing the culture medium with fresh complete medium
(RPMI 1640 supplemented with 10% FBS, glutamine, 0.1% gentamicin, and
1% antibiotic/antimycotic solution). Isolated KCs were identified after staining
using ED1 antibody (mouse antirat CD68, Serotec; 1:500 in blocking serum),
and a secondary antibody coupled to fluorescein isothiocyanate (Alexa Fluor
594 goat anti-mouse IgM [microchain]; Molecular Probes) as described
previously (Tukov et al., 2006). These techniques revealed a purity of � 90%
KCs after differential plating. In experiments using cocultures, HPCs (2.5 3
105/well) were added after differential plating of KCs, and the plates were
returned to the incubator.
Exposure of KC/HPC cocultures and KC monocultures to test agents. After
the cells were plated as described above, culture medium in KC/HPC cocultures or
KC monocultures was replaced with fresh, complete RPMI medium and allowed to
incubate overnight at 37�C. The medium was then removed and replaced with fresh
medium prior to treatment. Cell cultures were treated with LPS, RAN, FAM,
vehicle (culture medium), or a combination of LPS/RAN or LPS/FAM and
returned to the incubator for an additional 6 h, after which medium from each well
was collected and centrifuged at 9000 3 g for 5 min. A portion of the medium was
used for the determination of activity of ALT, and the remainder was stored at �80�C for future use.
Assessment of HPC toxicity in vitro. ALT activity in culture supernatants
was assayed as a measure of HPC injury using a diagnostic kit (ALT (GPT);
Ref: TR7 111-125, Infinity; Thermo Electron). Enzyme leakage into the
medium was expressed as a percentage of total intracellular ALT content,
which was determined in vehicle-treated control cells lysed with Triton X.
Experiments were considered valid only if the ALT activity in the medium of
vehicle-treated cells was less than 25% of total activity. TNF and PGE2 in
culture medium were determined as described above for serum.
Statistical analysis. Results are presented as means ± SEM. For these
studies, if the variances were homogenous, data were analyzed by Student’s
t-test or one-way or two-way ANOVA as appropriate. For data sets with
nonhomogenous variances, Kruskal-Wallis nonparametric ANOVA was used
or data were log-transformed prior to analysis using Student’s t-test or
ANOVA. Data points identified as statistical outliers using Grubb’s test were
not included in statistical analyses. For ANOVA, group means were compared
using Student-Newman-Keul post hoc test. Histopathology scores were
compared using Mann-Whitney rank sum test. For studies in vitro, two-way
repeated measures ANOVA was applied after appropriate data transformation,
and group comparisons were made using the Student Newman Keuhl post hoc
test. The criterion for significance was p < 0.05 for all studies.
FIG. 1. Time-dependent effect of LPS/RAN cotreatment on serum TNF-aconcentration. Rats fasted for 24 h were given 2.5 3 106 EU/kg LPS or its saline
vehicle iv. Two hours after LPS, either 30 mg RAN/kg or its vehicle was
administered iv. At 2, 3, 6, or 12 h after RAN administration, rats were killed, and
serum TNF-a concentration was determined using ELISA. Data are expressed as
mean ± SEM. n ¼ 4–13 rats/group. a, Significantly different from LPS/Veh
group at the same time; b, significantly different from Veh/Veh group at the same
time; c, significantly different from Veh/RAN group at the same time.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 269
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
Hours after drug1 2 3
ALT
(U/L
)
0
100
200
300
400
500
a,b,c
a,b,c
a,b,c
a,b,ca,b,c
a,b,ca,b,c
a,b,c
a,b,c
a,b,c
a,b,c
a,b,c
a,b,ca,b,c
a,b,c
b,c b,c
a,b
a,ba,b
a,b
a,ba,b
a,b,c
a,cTN
F (p
g/m
l)
0
1000
2000
3000Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
a
aa
aa a
BAIL
-1 b
eta
(pg/
ml)
0
100
200
300
400
500
600
700
aa
a a
a
C
IL-6
(pg/
ml)
0
20000
40000
60000
80000
100000D
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMVeh/LPSLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
Veh/VehVeh/RANVeh/FAMLPS/VehLPS/RANLPS/FAM
IFN
-gam
ma
(pg/
ml)
020
2000
4000
6000
8000
10000
12000
a
a,b
a
a
a
a a
G
PGE 2
(ng/
ml)
0
50
100
150
200
250
b
a a
H
F
PAI-1
(ng/
ml)
050
500
1000
1500
2000
2500
3000
aa
a a
a
a
IL-1
0 (p
g/m
l)
0
200
400
600
800
1000
1200
a aa a
a a
E
FIG. 2. Early liver injury and inflammatory mediator production after LPS/RAN cotreatment. Rats fasted for 24 h were given 2.5 3 106 EU/kg LPS or its
saline vehicle iv. Two hours after LPS, 30 mg RAN/kg, a pharmacologically equally efficacious dose of FAM (6 mg/kg) or its vehicle was administered, iv, and
rats were killed 1, 2, or 3 h later. Liver injury was assessed from the activity of ALT in serum (A). The concentrations of TNF-a (B), IL-1 beta (C), IL-6 (D), IL-10 (E),
PAI-1 (F), interferon-gamma (IFN) (G), and PGE2 (H) were determined as described in ‘‘Materials and Methods’’ section. Data are expressed as mean ± SEM.
n ¼ 4–9 rats/group. a, Significantly different from respective group in the absence of LPS; b, significantly different from respective group in the absence of drug;
c, significantly different from respective group in the presence of FAM.
270 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
RESULTS
Time-Course of Serum TNF Concentration in LPS/RAN-Cotreated Rats
At the time of RAN administration, the concentrations of
TNF in serum of rats treated 2 h earlier with LPS or saline were
32 ± 1 and 0.14 ± 0.05 ng/ml, respectively. This large LPS-
induced increase in TNF is consistent with previous reports
(Hewett et al., 1993). Neither vehicle nor RAN alone caused
a significant increase in serum TNF concentration over the 12-h
examination period (Fig. 1). Serum TNF concentration was
greater at all times in LPS-treated rats compared with vehicle-
treated rats. RAN cotreatment maintained plasma TNF
concentration at significantly greater values compared with
rats given LPS alone (Fig. 1). Serum TNF concentrations were
approximately twofold greater in LPS/RAN-cotreated animals
compared with LPS/veh-treated animals at 2 and 3 h after RAN
administration.
Early Inflammatory Mediator Release and Comparisonto FAM
Results from the experiment above suggested that RAN
enhanced the LPS-induced elevation of serum TNF around the
time that hepatocellular injury began to occur (i.e., 2–3 h
[Luyendyk et al., 2003, 2004b]). An additional experiment was
designed to (1) explore the effect of RAN on LPS-induced
TNF release before the onset of hepatocellular injury,
(2) compare the TNF-enhancing effects of RAN to those of
FAM, which does not interact with LPS to produce liver injury
in rats and does not have IADR liability in humans (Luyendyk
et al., 2006), and (3) elucidate the effects of RAN on other
inflammatory mediators. Rats were treated with a nontoxic
dose of LPS or its vehicle followed 2 h later by a non-
hepatotoxic dose of RAN, its vehicle, or a dose of FAM that is
pharmacologically equally efficacious to that of RAN (Luyendyk
et al., 2003). Times were chosen to encompass those before
(1 h) and during (2–3 h) the onset of hepatocellular injury. Of
the six treatments, only LPS/RAN was hepatotoxic, confirm-
ing previous results (Luyendyk et al., 2003, 2006). In this
group, serum ALT activity was unchanged 1 h after RAN
treatment, slightly elevated at 2 h, and more markedly elevated
by 3 h (Fig. 2A). Serum TNF concentration was not affected
by either RAN or FAM given alone (Fig. 2B). LPS exposure
elevated TNF concentration relative to control at all three of
these early times; RAN cotreatment enhanced this LPS effect
at all times, whereas FAM cotreatment did not influence it
(Fig. 2B). Similar effects of LPS-drug treatment were ob-
served for serum IL-1b, interferon-gamma, IL-6, IL-10, and
PAI-1 (Figs. 2C–G).
PGE2 is a product of cyclooxygenase-2 (COX-2), which is
expressed by cells during inflammatory responses. A previous
study indicated that COX-2 mRNA was upregulated in livers
of LPS/RAN-cotreated rats (Luyendyk et al., 2006). Further-
more, it has been reported that TNF increases production of
PGE2 and that PGE2 diminishes LPS-stimulated production of
TNF (Kreydiyyeh et al., 2007; Scales et al., 1989). Serum
PGE2 concentration was elevated slightly by either RAN or
FAM as well as by LPS (Fig. 2H). In LPS-treated rats,
cotreatment with either RAN or FAM enhanced serum PGE2
concentration compared to LPS treatment alone at all three
times.
Effect of PTX and Etan on LPS/RAN-Induced Liver Injury
In a preliminary efficacy study, the concentration of
circulating TNF was determined 2 h after administration of
LPS to rats pretreated with PTX or Etan. The LPS-induced
increase in TNF was markedly attenuated by pretreatment with
PTX (Fig. 3A). Similarly, Etan pretreatment significantly
reduced the increase in circulating TNF activity (Fig. 3B).
FIG. 3. Efficacy of PTX (A) and Etan (B) as inhibitors of TNF-a. Rats
fasted for 24 h received either PTX (100 mg/kg, iv) or its vehicle or Etan (8 mg/
kg, sc) or its vehicle 1 h before treatment with LPS (2.5 3 106 EU/kg, iv). Two
hours after LPS, rats were killed, and serum was collected. TNF-aconcentration was determined as described in ‘‘Materials and Methods’’
section. Data are expressed as mean ± SEM; n ¼ 4 rats/group. a, Significantly
different from the Veh/LPS group.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 271
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
LPS/RAN treatment increased serum ALT activity at 6 h
(Figs. 4A and 4B). Serum ALT activity was significantly less
in animals given either PTX (Fig. 4A) or Etan (Fig. 4B) prior to
LPS/RAN treatment. In preliminary studies, ALT activity was
not altered in Etan/veh/veh-treated animals compared with veh/
veh/veh-treated rats (data not shown). Similarly, we demon-
strated previously that the dose of PTX used for these studies
was not hepatotoxic in rats (Hewett et al., 1993). These
findings were supported by histologic examination of liver
sections. Liver from veh/veh/veh-treated animals appeared
normal (Fig. 5A). Sections from LPS/RAN-cotreated rats taken
6 h after administration of RAN had acute, multifocal,
midzonal hepatocellular necrosis (Fig. 5B). These lesions have
been described in detail previously (Luyendyk et al., 2003).
In livers from animals that received PTX (Fig. 5C) or Etan
(Fig. 5D) prior to LPS/RAN cotreatment, the size and
frequency of the necrotic foci were significantly reduced
compared with the LPS/RAN-treated rats.
Effect of PTX or Etan on LPS/RAN-Induced Increasesin Serum IL-1b and IL-6 Concentrations
Cotreatment of rats with LPS/RAN significantly increased
circulating concentrations of IL-1b and IL-6 compared with
vehicle-treated rats (Fig. 6). Pretreatment with PTX reduced the
LPS/RAN-induced increase in IL-1b and IL-6 (Figs. 6A and
6B). Etan reduced the LPS/RAN-induced increase in circulat-
ing IL-1b (Fig. 6C) but did not significantly affect IL-6 con-
centration, although there was a trend toward reduced
concentration (Fig. 6D).
Effect of PTX or Etan on LPS/RAN-Induced Changesin the Hemostatic System
Increased TAT concentration in plasma is a biomarker of
coagulation system activation. PAI-1 is the major endogenous
inhibitor of fibrinolysis, thus an increase in its concentration in
plasma suggests impaired fibrinolysis. Previous studies dem-
onstrated that both TAT and PAI-1 concentrations were
enhanced early after LPS/RAN treatment (Fig. 2F; Luyendyk
et al., 2004a,b). Confirming these results, LPS/RAN treatment
increased plasma TAT concentration and total PAI-1 concen-
tration in serum (Fig. 7). These increases were significantly
attenuated by either PTX or Etan pretreatment (Fig. 7).
FIG. 4. Effect of PTX (A) or Etan (B) on the LPS/RAN-induced elevation
in serum ALT activity. Rats fasted for 24 h received either PTX (100 mg/kg, iv)
or its vehicle or Etan (8 mg/kg, sc) or its vehicle 1 h before treatment with LPS
(2.5 3 106 EU/kg, iv) or its vehicle. Two hours after LPS, either RAN (30 mg/kg,
iv) or its vehicle was administered. Hepatocellular injury was assessed 6 h after
RAN as increased serum ALT activity. Data are expressed as mean ± SEM; n ¼4–6 rats/group; a, significantly different from the veh/veh/veh group; b,
significantly different from the veh/LPS/RAN group.
FIG. 5. Representative photomicrographs of livers after LPS/RAN
cotreatment. Rats were treated as described in the legend to Fig. 4. At 6 h
after RAN administration, rats were killed and livers were removed and fixed in
10% neutral-buffered formalin, stained with H&E, and examined by light
microscopy. (A) veh/veh/veh, (B) veh/LPS/RAN, (C) PTX/LPS/RAN, and
(D) Etan/LPS/RAN-treated rats. The photomicrographs were taken at 1003
magnification. Acute, multifocal, midzonal hepatocellular necrosis (arrows)
developed in LPS/RAN-cotreated rats. Reduction in the size and frequency of
necrotic lesions was observed in PTX- and Etan-pretreated rats.
272 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
Liver sections from veh/veh/veh-treated animals showed
very little fibrin deposition, with some staining of the larger ves-
sels due to fibrin deposition during animal sacrifice (Luyendyk
et al., 2003) (Fig. 8A). Fibrin deposition in liver tissues from
veh/LPS/RAN-treated rats (Fig. 8B) was significantly in-
creased compared with control. Pretreatment with either PTX
(Fig. 8C) or Etan (Fig. 8D) attenuated the increase in fibrin
deposition caused by veh/LPS/RAN treatment. Morphometric
analysis confirmed this result (Figs. 8E and 8F).
Effect of PTX or Etan on LPS/RAN-Induced Hepatic PMNAccumulation and Serum MIP-2 Concentration
Previous results suggested that PMNs play a causal role in
LPS/RAN-induced hepatocellular injury (Luyendyk et al.,2005). To test the possibility that TNF mediates liver injury
after LPS/RAN exposure by promoting PMN accumulation in
the liver, the effects of PTX or Etan pretreatment on the LPS/
RAN-induced increase in liver PMNs were evaluated.
Quantitative analysis of liver sections immunostained for
PMNs revealed that PMN accumulation occurred within 6 h
in livers from animals treated with LPS/RAN. PMN numbers
in the livers of LPS/RAN-treated animals were unaffected by
either PTX or Etan (Figs. 9A and 9C). PMN distribution was
panlobular in livers of LPS/RAN-cotreated rats, although foci
of PMNs were noted in lesioned areas.
MIP-2 is a chemokine that can cause transmigration of
PMNs from vessels and PMN activation. The serum concentra-
tion of MIP-2 was significantly increased in rats cotreated with
LPS and RAN. This increase was attenuated by pretreatment
with either PTX or Etan (Figs. 9B and 9D).
Mediator Production In Vitro
KCs are major producers of the inflammatory mediators
altered by LPS/RAN cotreatment in vivo. Accordingly, release
of mediators was evaluated in vitro in KC/HPC cocultures at
noncytotoxic drug and LPS concentrations. Concentrations of
RAN were chosen to approximate the maximum concentration
likely to occur in plasma in the study in vivo above (i.e., approx-
imately 100lM) and one larger, but noncytotoxic concentration
(i.e., 500lM). Neither LPS, RAN, FAM nor combinations of
LPS and the drugs caused release of ALT in KC/HPC
cocultures within 6 h of incubation (Figs. 10A and 10B). LPS
increased TNF release (Figs. 10C and 10D). RAN enhanced
slightly the LPS-induced TNF release, but only at the larger
RAN concentration, whereas both FAM concentrations led to
a modest enhancement. Unlike the results in KC/HPC
cocultures, RAN was without effect on LPS-stimulated
TNF release when it was added to KCs cultured by themselves
(Fig. 11A), suggesting that the presence of HPCs modified
the ability of RAN to alter this response. At the larger
FIG. 6. Effect of TNF-a inhibition on proinflammatory cytokine release after LPS/RAN cotreatment. Rats were treated as described in the legend to Fig. 4. At
6 h after RAN administration, rats were killed, and serum IL-1b (A and C) and IL-6 (B and D) concentrations were determined by ELISA. Data are expressed as
mean ± SEM; n ¼ 5–6 rats/group; a, Significantly different from the veh/veh/veh group; b, significantly different from the veh/LPS/RAN group.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 273
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
concentration, FAM by itself elicited TNF release in KC
cultures (Fig. 11B).
MIP-2, the release of which was selectively enhanced by
RAN in LPS-cotreated rats (Luyendyk et al., 2006), was un-
affected by RAN or FAM cotreatment in KC/HPC cocultures
(Figs. 12A and 12B). Interestingly, PGE2 release in vitro was
largely unaffected by LPS but enhanced markedly by RAN
(Fig. 12C). In contrast, FAM caused only a slight enhancement
(Fig. 12D).
DISCUSSION
TNF is an important proinflammatory cytokine involved in
normal physiological immune and inflammatory processes.
A causative role for TNF in promoting liver injury has been
identified for inflammation interaction with a number of
xenobiotic agents (Barton et al., 2001; Endo et al., 1999;
Yee et al., 2002) as well as for large, hepatotoxic doses of LPS
(Hewett et al., 1993). The present work explored the role of
TNF in an animal model of RAN-induced idiosyncratic liver
injury. LPS exposure induces a large and rapid increase in
plasma TNF concentration (Hewett et al., 1993). In the current
study, despite increased plasma TNF concentration, rats given
a small dose of LPS did not develop liver injury by 3 h. This
indicates that an increase in TNF of this magnitude and
duration is not sufficient to cause hepatocellular damage.
Despite this insufficiency, this large increase in TNF con-
centration could be critical for the genesis of hepatocellular
injury when combined with other, independent effects of RAN
cotreatment. Indeed, RAN-cotreated rats sustained greater
plasma TNF levels compared with rats given LPS alone and
developed hepatotoxicity. In contrast, FAM at an equally
efficacious dose neither enhanced TNF production nor
synergized with LPS to produce liver injury. Thus, the increase
in LPS-stimulated TNF production distinguished a drug that
causes human IADRs from one that does not. This also raises
the possibility that the effect of RAN on LPS-induced TNF
production could contribute to hepatotoxicity in LPS/RAN-
treated rats.
To determine the role of TNF in LPS/RAN-induced
hepatotoxicity, we used two approaches. PTX is a methylxanthine
that inhibits the synthesis of TNF (Barton et al., 2001; Dezube
et al., 1993; Yee et al., 2003). However, PTX also has several
other pharmacological effects (Banfi et al., 2004). Accordingly,
we also used Etan, a dimeric fusion protein that contains a soluble
TNF receptor capable of selectively neutralizing TNF. PTX and
Etan significantly reduced serum TNF concentration and activity,
respectively, and were both effective in reducing hepatocellular
FIG. 7. Effect of PTX or Etan on plasma TAT and serum PAI-1 concentrations after LPS/RAN cotreatment. Rats were treated as described in the legend to
Fig. 4. At 6 h after RAN administration, they were killed, and plasma TAT (A and C) and serum PAI-1 (B and D) concentrations were determined by ELISA. Data
are expressed as mean ± SEM; n ¼ 5–6 rats/group; a, significantly different from the veh/veh/veh group; b, significantly different from the veh/LPS/RAN group.
274 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
injury in LPS/RAN-cotreated rats. These results indicate that
TNF is involved in LPS/RAN-induced liver injury.
Although this enhancement of TNF concentration by RAN
was small compared with the pronounced increase caused by
LPS exposure, the prolongation of TNF levels above
a particular threshold could have pathogenic importance since
it occurred prior to injury in LPS/RAN-treated rats. The
importance of the RAN-induced potentiation of LPS-induced
TNF release is difficult to test. We have shown previously that
LPS given alone at the dose used in these studies does not
cause liver injury up to 24 h (Luyendyk et al., 2003).
Accordingly, LPS-induced TNF is insufficient to cause liver
damage. TNF induces the expression of several cytokines
including IL-1b, IL-6, IL-8, MIP-2, as well as other gene
products capable of damaging the liver directly or in concert
with TNF (Aggarwal and Natarajan, 1996; Locksley et al.,2001). For example, IL-1b and IL-6 display partially over-
lapping activities with TNF (Eigler et al., 1997). TNF and
FIG. 8. Effects of PTX or Etan on hepatic fibrin deposition after LPS/RAN cotreatment. Rats were treated as described in the legend to Fig. 4. Livers were
removed 6 h after RAN administration and processed for fibrin immunohistochemistry as described in ‘‘Materials and Methods’’ section. Fibrin appears as white on
a gray background. Representative liver section from (A) a veh/veh/veh-treated rat showing minimal fibrin staining, (B) a veh/LPS/RAN-treated rat with sinusoidal
fibrin deposits in both periportal and centrilobular regions and (C) a PTX/LPS/RAN-treated animal, and (D) an Etan/LPS/RAN-treated animals showing reductions in
fibrin staining compared with those from veh/LPS/RAN-treated rats. Fibrin deposition was quantified as described in ‘‘Materials and Methods’’ section. (E and F).
Data are expressed as mean ± SEM; n¼ 5–6 rats/group. a, significantly different from the veh/veh/veh group; b, significantly different from the veh/LPS/RAN group.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 275
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
IL-1b activate PMNs (Klebanoff et al., 1986), induce their own
production, stimulate production of other inflammatory medi-
ators, and antagonize anti-inflammatory cytokines such as
IL-10 (Ji et al., 2004). In the present study, LPS induction of
IL-1b and IL-6 was enhanced by RAN cotreatment but not by
cotreatment with FAM. The LPS/RAN-induced increase in
these cytokines was inhibited by TNF neutralization, suggest-
ing that the RAN-dependent increase in TNF is required for
enhanced expression of inflammatory genes in LPS/RAN-
treated rats (Figs. 2C and 2D and 6B and 6C).
In addition to enhancing cytokine expression, TNF–
dependent gene expression might contribute to LPS/RAN-
induced liver injury by enhancing coagulation. RAN cotreatment
enhanced LPS-induced coagulation prior to liver injury, and
anticoagulants reduced liver damage in LPS/RAN-treated rats
(Luyendyk et al., 2004a). TNF and other cytokines, including
IL-1b and IL-6, induce tissue factor expression on endothelial
cells and monocytes/macrophages in vitro (Neumann et al.,1997). Tissue factor is generally accepted as the pivotal
initiator of coagulation system activation in endotoxemia and
bacteremia (de Jonge et al., 2000; Levi et al., 1994; Pawlinski
et al., 2004; Taylor et al., 1991). In addition, in the presence of
PMNs, TNF causes sinusoidal endothelial cell (SEC) damage
in vitro (Smedly et al., 1986; Takei et al., 1995), which can
activate the coagulation system (Colman et al., 1994). Indeed,
PTX and Etan inhibited coagulation activation in LPS/RAN-
treated rats, as marked by reduced plasma TAT concentration
(Fig. 7). This result indicates that TNF contributes to LPS/
RAN-induced liver injury at least in part by enhancing
activation of the coagulation system.
Activation of the coagulation system in LPS/RAN-treated rats
results in the formation of fibrin clots in liver sinusoids, and
prevention of fibrin deposition was associated with reduced
hepatocellular injury (Luyendyk et al., 2004a). Neutralization of
TNF reduced LPS/RAN-induced liver fibrin deposition (Fig. 8).
This reduction could be a consequence of reduced coagulation
and/or increased fibrinolysis. TNF has been shown to play a role
in the induction of PAI-1 in the liver and plasma of LPS-treated
mice (Fearns and Loskutoff, 1997). Like TNF, the concentration
of PAI-1 in the plasma was enhanced by RAN cotreatment
(Fig. 2F, Luyendyk et al., 2006), and TNF neutralization
inhibited the induction of PAI-1 in LPS/RAN-treated rats
(Figs. 7B and 7D). The cellular source and mechanism of
enhanced PAI-1 expression in livers of LPS/RAN-cotreated rats
is not known. TNF and IL-1 can stimulate the expression and
release of PAI-1 by endothelial cells (Schleef et al., 1988).
Indeed, RAN cotreatment enhanced the LPS-induced alteration
of SEC function (Luyendyk et al., 2004a and b). Taken together,
these results suggest that TNF enhancement of PAI-1 production
promotes fibrin deposition and injury in this animal model.
FIG. 9. Effect of PTX or Etan on serum MIP-2 concentration and the accumulation of PMNs in liver after LPS/RAN cotreatment. Rats were treated as
described in the legend to Fig. 4. Serum MIP-2 concentration (B and D) was evaluated 6 h after RAN treatment. Liver PMN accumulation (A and C) was evaluated
in 10–20 randomly selected 4003 fields 6 h after RAN administration as described under ‘‘Materials and Methods’’ section. Data are expressed as mean ± SEM;
n ¼ 5–6 rats/group. a, Significantly different from the veh/veh/veh group. b, Significantly different from the veh/LPS/RAN group.
276 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
Coagulation-dependent liver damage in LPS/RAN-cotreated
rats could activate inflammatory cells such as PMNs. Depletion
of circulating PMNs attenuated LPS/RAN-induced liver injury,
suggesting their importance in the pathogenesis (Luyendyk
et al., 2005). In models of PMN-dependent liver injury, PMNs
not only accumulate in liver sinusoids but must also trans-
migrate through sinusoidal endothelium and become activated
in close proximity to HPCs to cause injury to these cells
(Jaeschke, 2006). Interestingly, anticoagulation did not inhibit
PMN accumulation in LPS/RAN-treated rats, but it did reduce
plasma concentration of the PMN chemokine, MIP-2
(Luyendyk et al., 2006), which can activate PMNs and elicit
the release of toxic products (Biedermann et al., 2000; Wang
and Thorlacius, 2005). TNF can prompt the accumulation of
PMNs in tissues by activating endothelial cells (Bradham et al.,1998; Vassalli, 1992) and can also prime PMNs for activation
FIG. 10. Release of ALT and TNF-a from HPC/KC cocultures exposed to LPS and RAN or FAM. HPCs and KCs were isolated and cocultured as described
in ‘‘Materials and Method’’ section. LPS (10 ng/ml) and drug at the indicated concentrations were added at the same time, and the medium was collected after 6 h.
ALT activity and TNF-a concentration were measured in the cell-free supernatant. ALT activity (A and B) in the medium is expressed as a percent of the total ALT
activity. TNF-a concentration (C and D) was determined by ELISA. n ¼ 4–5 isolations from different rats. a Significantly different from vehicle-treated group at
the same concentration of drug; b, significantly different from corresponding vehicle-treated group in the absence of drug.
FIG. 11. Release of TNF-a from KCs exposed to LPS and RAN or FAM. KCs were isolated and cultured as described in ‘‘Materials and Method’’ section.
LPS (10 ng/ml) and drug at the indicated concentrations were added at the same time, and medium was collected after 6 h. TNF-a concentration was measured in
the cell-free supernatant. n ¼ 3–4 isolations from different rats. a, Significantly different from corresponding vehicle-treated group at the same concentration of
drug; b significantly different from vehicle-treated group in the absence of drug.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 277
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
(Kushimoto et al., 1996; Nagaki et al., 1991; Vassalli, 1992).
Analogous to the protection afforded by anticoagulants in this
model, inhibition of TNF did not impact hepatic PMN
accumulation but did reduce serum MIP-2 concentration
(Fig. 9). The observations that TNF inhibition reduced the
concentration of MIP-2 in plasma and attenuated coagulation
system activation and that inhibition of coagulation reduced
MIP-2 concentration (Luyendyk et al., 2006) are consistent
with the hypothesis that TNF activation of coagulation induces
the expression of MIP-2, which could activate PMNs
accumulated in the liver.
Recently, we reported that inflammatory mediator expres-
sion elicited by xenobiotic/LPS coexposure in rats could be
recapitulated using a KC/HPC coculture system (Tukov et al.,2006). In the present study, RAN enhanced TNF release in the
presence of LPS in KC/HPC cocultures but only at a large
RAN concentration that is unlikely to be achieved in the LPS/
RAN animal model (Fig. 10). FAM had a similar effect. Thus,
this in vitro cytokine system failed to distinguish a drug that
causes IADRs in humans from one that does not. The MIP-2
response in vitro also differed from the response in rats
(Luyendyk et al., 2006, and Fig. 12). The disparity between
these in vitro and in vivo responses suggests that the RAN-
induced enhancement of TNF and MIP-2 production seen
in vivo likely occurs by indirect mechanisms, perhaps involving
other cell types.
LPS/RAN treatment of rats selectively increased COX-2
mRNA expression, whereas LPS/FAM treatment did not
(Luyendyk et al., 2006). Interestingly, treatment with RAN
produced a similar increase in PGE2 compared with treatment
FIG. 12. Release of MIP-2 and PGE2 from HPC/KC cocultures exposed to LPS and RAN or FAM. HPCs and KCs were isolated and cocultured as described
in ‘‘Materials and Method’’ section. LPS (10 ng/ml) and drug at the indicated concentrations were added at the same time, medium was collected after 6 h, and
MIP-2 (A and B) and PGE2 (C and D) were measured in the cell-free supernatant. n ¼ 3–5 isolations from different rats. a, Significantly different from vehicle-
treated group at the same concentration of drug; b, significantly different from corresponding vehicle-treated group in the absence of drug.
FIG. 13. Working hypothesis for the pathogenesis of LPS/RAN-induced hepatotoxicity. Previous studies (see ‘‘Discussion’’ and Luyendyk et al. 2004a and b;
2005) have implicated the hemostatic system and neutrophils (PMNs) as critical players in the hepatotoxic interaction between LPS and RAN. The present results
suggest that TNF-a is important as a proximal mediator in the chain of events leading to hepatocellular injury.
278 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
with FAM, irrespective of LPS cotreatment (Fig. 2H). In KC/
HPC cocultures, RAN prompted the release of PGE2, whereas
FAM had little effect (Figs. 12C and 12D). These drug effects
were not influenced by LPS treatment of the cells. Thus,
increased PGE2 release did distinguish the two drugs in vitro,
whereas it failed to do so in vivo.
In summary, RAN cotreatment potentiated the LPS-medi-
ated production of TNF at a time before the onset of liver injury
in LPS/RAN-cotreated rats. Inhibition of TNF biosynthesis or
signaling significantly attenuated LPS/RAN-induced hepato-
cellular injury. Interference with TNF also reduced the serum
concentrations of other proinflammatory cytokines and chemo-
kines in LPS/RAN-treated rats. In addition, TNF inhibition
decreased coagulation system activation, PAI-1 production,
and fibrin deposition. Together, these results suggest that TNF
contributes to LPS/RAN-induced liver injury by activating the
hemostatic system, inducing inflammatory cytokines, and
possibly by contributing to PMN activation (Fig. 13).
FUNDING
Grant DK061315 from the National Institutes of Health;
Training Grant 2 T32 ES07255 from the National Institute of
Environmental Health Sciences to F.F.T. and J.P.L.; the
Society of Toxicology’s Colgate Palmolive Postdoctoral
Fellowship in In Vitro Toxicology to F.F.T.; and Society of
Toxicology’s Novartis Predoctoral Fellowship to J.P.L.
ACKNOWLEDGMENTS
The authors are grateful to Theresa Eagle and Sandra Newport for technical
assistance.
REFERENCES
Aderka, D., Engelmann, H., Wysenbeek, A. J., and Levo, Y. (1992). The
possible role of tumor necrosis factor (TNF) and its natural inhibitors, the
soluble-TNF receptors, in autoimmune diseases. Isr. J. Med. Sci. 28,
126–130.
Aggarwal, B. B., and Natarajan, K. (1996). Tumor necrosis factors: Develop-
ments during the last decade. Eur. Cytokine Netw. 7, 93–124.
Ament, P. W., Roth, J. D., and Fox, C. J. (1994). Famotidine-induced mixed
hepatocellular jaundice. Ann. Pharmacother. 28, 40–42.
Banfi, C., Sironi, L., De Simoni, G., Gelosa, P., Barcella, S., Perego, C.,
Gianazza, E., Guerrini, U., Tremoli, E., and Mussoni, L. (2004). Pentoxifyl-
line prevents spontaneous brain ischemia in stroke-prone rats. J. Pharmacol.
Exp. Ther. 310, 890–895.
Barton, C. C., Barton, E. X., Ganey, P. E., Kunkel, S. L., and Roth, R. A.
(2001). Bacterial lipopolysaccharide enhances aflatoxin B1 hepatotoxicity in
rats by a mechanism that depends on tumor necrosis factor alpha.
Hepatology 33, 66–73.
Beutler, B., and Cerami, A. (1988). Tumor necrosis, cachexia, shock, and
inflammation: A common mediator. Annu. Rev. Biochem. 57, 505–518.
Biedermann, T., Kneilling, M., Mailhammer, R., Maier, K., Sander, C. A.,
Kollias, G., Kunkel, S. L., Hultner, L., and Rocken, M. (2000). Mast cells
control neutrophil recruitment during T cell-mediated delayed-type hyper-
sensitivity reactions through tumor necrosis factor and macrophage inflam-
matory protein 2. J. Exp. Med. 192, 1441–1452.
Bradham, C. A., Plumpe, J., Manns, M. P., Brenner, D. A., and Trautwein, C.
(1998). Mechanisms of hepatic toxicity. I. TNF-induced liver injury. Am.
J. Physiol. 275, G387–G392.
Colman, R. W., Marder, V. J., Salzman, E. W., and Hirsh, J. (1994). Overview
of hemostasis. In Hemostasis and Thrombosis (R. W. Colman, J. Hirsh,
V. J. Marder, and E. W. Salzman, Eds.), p. 3. Lippincott, Philadelphia.
Copple, B. L., Banes, A., Ganey, P. E., and Roth, R. A. (2002). Endothelial cell
injury and fibrin deposition in rat liver after monocrotaline exposure.
Toxicol. Sci. 65, 309–318.
de Jonge, E., Dekkers, P. E. P., Creasey, A. A., Hack, C. E., Paulson, S. K.,
Karim, A., Kesecioglu, J., Levi, M., van Deventer, S. J. H., and van der
Poll, T. (2000). Tissue factor pathway inhibitor dose-dependently inhibits
coagulation activation without influencing the fibrinolytic and cytokine
response during human endotoxemia. Blood 95, 1124–1129.
Dezube, B. J., Sherman, M. L., Fridovich-Keil, J. L., Allen-Ryan, J., and
Pardee, A. B. (1993). Down-regulation of tumor necrosis factor expression
by pentoxifylline in cancer patients: A pilot study. Cancer Immunol.
Immunother. 36, 57–60.
Doherty, G. M., Jensen, J. C., Alexander, H. R., Buresh, C. M., and
Norton, J. A. (1991). Pentoxifylline suppression of tumor necrosis factor
gene transcription. Surgery 110, 192–198.
Eigler, A., Sinha, B., Hartmann, G., and Endres, S. (1997). Taming TNF:
Strategies to restrain this proinflammatory cytokine. Immunol. Today 18,
487–492.
Endo, Y., Shibazaki, M., Yamaguchi, K., Kai, K., Sugawara, S., Takada, H.,
Kikuchi, H., and Kumagai, K. (1999). Enhancement by galactosamine of
lipopolysaccharide (LPS)-induced tumour necrosis factor production and
lethality: Its suppression by LPS pretreatment. Br. J. Pharmacol. 128, 5–12.
Eskandari, M. K., Nguyen, D. T., Kunkel, S. L., and Remick, D. G. (1990).
WEHI 164 subclone 13 assay for TNF: Sensitivity, specificity, and
reliability. Immunol. Investig. 19, 69–79.
Espevik, T., and Nissen-Meyer, J. (1986). A highly sensitive cell line, WEHI
164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from
human monocytes. J. Immunol. Methods 95, 99–105.
Fearns, C., and Loskutoff, D. J. (1997). Induction of plasminogen activator
inhibitor type 1 gene expression in murine liver by lipopolysaccharide.
Cellular localization and role of endogenous tumor necrosis factor-a. Am.
J. Pathol. 150, 579–590.
Ganey, P. E., and Roth, R. A. (2001). Concurrent inflammation as a determinant
of susceptibility to toxicity from xenobiotic agents. Toxicology 169, 195–208.
Geier, A., Dietrich, C. G., Voigt, S., Kim, S. K., Gerloff, T., Kullak-
Ublick, G. A., Lorenzen, J., Matern, S., and Gartung, C. (2003). Effects of
proinflammatory cytokines on rat organic anion transporters during toxic
liver injury and cholestasis. Hepatology 38, 345–354.
Hashimoto, F., Davis, R. L., and Egli, D. (1994). Hepatitis following treatments
with famotidine and then cimetidine. Ann. Pharmacother. 28, 37–39.
Hewett, J. A., Jean, P. A., Kunkel, S. L., and Roth, R. A. (1993). Relationship
between tumor necrosis factor- and neutrophils in endotoxin-induced liver
injury. Am. J. Physiol. 265, G1011–G1015.
Hewett, J. A., Schultze, A. E., Van Cise, S., and Roth, R. A. (1992). Neutrophil
depletion protects against liver injury from bacterial endotoxin. Lab.
Investig. 66, 347–361.
Jaeschke, H. (2006). Mechanisms of liver injury. II. Mechanisms of neutrophil-
induced liver cell injury during hepatic ischemia-reperfusion and other acute
inflammatory conditions. Am. J. Physiol. Gastrointest. Liver Physiol. 290,
G1083–G1088.
ROLE OF TNF-a IN LPS/RAN-INDUCED LIVER INJURY 279
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from
Ji, Q., Zhang, L., Jia, H., and Xu, J. (2004). Pentoxifylline inhibits endotoxin-
induced NF-kappa B activation and associated production of proinflamma-
tory cytokines. Ann. Clin. Lab. Sci. 34, 427–436.
Jimenez-Saenz, M., Arguelles-Arias, F., Herrerias-Gutierrez, J. M., and Duran-
Quintana, J. A. (2000). Acute cholestatic hepatitis in a child treated with
famotidine. Am. J. Gastroenterol. 95, 3665–3666.
Klebanoff, S. J., Vadas, M. A., Harlan, J. M., Sparks, L. H., Gamble, J. R.,
Agosti, J. M., and Waltersdorph, A. M. (1986). Stimulation of neutrophils by
tumor necrosis factor. J. Immunol. 136, 4220–4225.
Kreydiyyeh, S. I., Riman, S., Serhan, M., and Kassardjian, A. (2007). TNF-
alpha modulates hepatic Naþ-Kþ ATPase activity via PGE2 and EP2
receptors. Prostaglandins Other Lipid Mediat. 83, 295–303.
Kushimoto, S., Okajima, K., Uchiba, M., Murakami, K., Okabe, H., and
Takatsuki, K. (1996). Role of granulocyte elastase in ischemia/reperfusion
injury of rat liver. Crit. Care Med. 24, 1908–1912.
Levi, M., ten Cate, M., Bauer, K. A., van der Poll, T., Edgington, T. S.,
Buller, H. R., van Deventer, S. J., Hack, C. E., ten Cate, J. W., and
Rosenberg, R. D. (1994). Inhibition of endotoxin-induced activation of
coagulation and fibrinolysis by pentoxifylline or by a monoclonal anti-tissue
factor antibody in chimpanzees. J. Clin. Investig. 93, 114–120.
Lin, J. H. (1991). Pharmacokinetic and pharmacodynamic properties of
histamine H2-receptor antagonists: Relationship between intrinsic potency
and effective plasma concentrations. Clin. Pharmacokinet. 20, 218–236.
Locksley, R. M., Killeen, N., and Lenardo, M. J. (2001). The TNF and TNF
receptor superfamilies: Integrating mammalian biology. Cell 104, 487–501.
Luyendyk, J. P., Lehman-McKeeman, L. D., Nelson, D. M., Bhaskaran, V. M.,
Reilly, T. P., Car, B. D., Cantor, G. H., Maddox, J. F., Ganey, P. E., and
Roth, R. A. (2006). Unique gene expression and hepatocellular injury in the
lipopolysaccharide-ranitidine drug idiosyncrasy rat model: Comparison with
famotidine. Toxicol. Sci. 90, 569–585.
Luyendyk, J. P., Maddox, J. F., Cosma, G. N., Ganey, P. E., Cockerell, G. L.,
and Roth, R. A. (2003). Ranitidine treatment during a modest inflammatory
response precipitates idiosyncrasy-like liver injury in rats. J. Pharmacol.
Exp. Ther. 307, 9–16.
Luyendyk, J. P., Maddox, J. F., Green, C. D., Ganey, P. E., and Roth, R. A.
(2004a). Role of hepatic fibrin in idiosyncrasy-like liver injury from
lipopolysaccharide-ranitidine coexposure in rats. Hepatology 40, 1342–1351.
Luyendyk, J. P., Mattes, W. B., Burgoon, L. D., Zacharewski, T. R.,
Maddox, J. F., Cosma, G. N., Ganey, P. E., and Roth, R. A. (2004b). Gene
expression analysis points to hemostasis in livers of rats cotreated with
lipopolysaccharide and ranitidine. Toxicol. Sci. 80, 203–213.
Luyendyk, J. P., Shaw, P. J., Green, C. D., Maddox, J. F., Ganey, P. E., and
Roth, R. A. (2005). Coagulation-mediated hypoxia and neutrophil-dependent
hepatic injury in rats given lipopolysaccharide and ranitidine. J. Pharmacol.
Exp. Ther. 314, 1023–1031.
Nagaki, M., Muto, Y., Ohnishi, H., and Moriwaki, H. (1991). Significance of
tumor necrosis factor (TNF) and interleukin-1 (IL-1) in the pathogenesis of
fulminant hepatitis: Possible involvement of serine protease in TNF-
mediated liver injury. Gastroenterol. Jpn. 26, 448–455.
Neumann, J. F., Ott, I., Marx, N., Luther, T., Kenngott, S., and Gawaz, M.
(1997). Effect of human recombinant interleukin-6 and interleukin-8 on
monocyte procoagulant activity. Arterioscler. Thromb. Vasc. Biol. 17,
3399–3405.
Pawlinski, R., Pedersen, B., Schabbauer, G., Tencati, M., Holscher, T.,
Boisvert, W., Andrade-Gordon, P., Frank, R. D., and Mackman, N. (2004).
Role of tissue factor and protease-activated receptors in a mouse model of
endotoxemia. Blood 103, 1342–1347.
Ribeiro, J. M., Lucas, M., Baptista, A., and Victorino, R.-R. M. (2000). Fatal
hepatitis associated with ranitidine. Am. J. Gastroenterol. 95, 559–560.
Roth, R. A., Harkema, J. R., Pestka, J. P., and Ganey, P. E. (1997). Is exposure
to bacterial endotoxin a determinant of susceptibility to intoxication from
xenobiotic agents? Toxicol. Appl. Pharmacol. 147, 300–311.
Scales, W. E., Chensue, S. W., Otterness, I., and Kunkel, S. L. (1989).
Regulation of monokine gene expression: Prostaglandin E2 suppresses tumor
necrosis factor but not interleukin-1 alpha or beta-mRNA and cell-associated
bioactivity. J. Leukoc. Biol. 45, 416–421.
Scarpignato, C., Tramacere, R., and Zappia, L. (1987). Antisecretory and
antiulcer effect of H2-receptor antagonist famotidine in the rat: Comparison
with ranitidine. Br. J. Pharmacol. 92, 153–159.
Schandene, L., Vandenbussche, P., Crusiaux, A., Alegre, M. L.,
Abramowicz, D., Dupont, E., Content, J., and Goldman, M. (1992).
Differential effects of pentoxifylline on the production of tumour necrosis
factor-alpha (TNF-alpha) and interleukin-6 (IL-6) by monocytes and T cells.
Immunology 76, 30–34.
Schleef, R. R., Bevilaqua, M. P., Sawdey, M., Gimbrone, M. A., and
Loskutoff, D. J. (1988). Cytokine activation of vascular endothelium. Effects
on tissue-type plasminogen activator and type 1 plasminogen activator
inhibitor. J. Biol. Chem. 263, 5797–5803.
Smedly, L. A., Tonnesen, M. G., Sandhaus, R. A., Haslett, C., Guthrie, L. A.,
Johnston, R. B., Jr, Henson, P. M., and Worthen, G. S. (1986). Neutrophil-
mediated injury to endothelial cells. Enhancement by endotoxin and essential
role of neutrophil elastase. J. Clin. Investig. 77, 1233–1243.
Sneed, R. A., Buchweitz, J. P., Jean, P. A., and Ganey, P. E. (2000).
Pentoxifylline attenuates bacterial lipopolysaccharide-induced enhancement
of allyl alcohol hepatotoxicity. Toxicol. Sci. 56, 203–210.
Takei, Y., Kawano, S., Nishimura, Y., Goto, M., Nagai, H., Chen, S. S.,
Omae, A., Fusamoto, H., Kamada, T., Ikeda, K., et al. (1995). Apoptosis:
A new mechanism of endothelial and Kupffer cell killing. J. Gastroenterol.
Hepatol. 10(Suppl. 1), S65–S67.
Taylor, F. B., Jr, Chang, A., Ruf, W., et al. (1991). Lethal E. coli septic shock is
prevented by blocking tissue factor with monoclonal antibody. Circ. Shock
33, 127–134.
Tukov, F. F., Maddox, J. F., Amacher, D. E., Bobrowski, W. F., Roth, R. A.,
and Ganey, P. E. (2006). Modeling inflammation-drug interactions in vitro:
A rat Kupffer cell-hepatocyte coculture system. Toxicol. In Vitro 20,
1488–1499.
van der Poll, T., Coyle, S. M., Levi, M., Jansen, P. M., Dentener, M.,
Barbosa, K., Buurman, W. A., Hack, C. E., ten Cate, J. W., Agosti, J. M.,
et al. (1997). Effect of a recombinant dimeric tumor necrosis factor receptor
on inflammatory responses to intravenous endotoxin in normal humans.
Blood 89, 3727–3734.
Vassalli, P. (1992). The pathophysiology of tumor necrosis factors. Annu. Rev.
Immunol. 10, 411–452.
Vial, T., Goubier, C., Bergeret, A., Cabrera, F., Evreux, J.-C., and Descotes, J.
(1991). Side effects of ranitidine. Drug Saf. 6, 94–117.
Wang, Y., and Thorlacius, H. (2005). Mast cell-derived tumour necrosis factor-
alpha mediates macrophage inflammatory protein-2-induced recruitment of
neutrophils in mice. Br. J. Pharmacol. 145, 1062–1068.
Yee, S. B., Ganey, P. E., and Roth, R. A. (2002). Temporal relationships in
the augmentation of monocrotaline hepatoxicity by bacterial endotoxin.
J. Toxicol. Environ. Health A 65, 961–976.
Yee, S. B., Ganey, P. E., and Roth, R. A. (2003). The role of Kupffer cells and
TNF-alpha in monocrotaline and bacterial lipopolysaccharide-induced liver
injury. Toxicol. Sci. 71, 124–132.
280 TUKOV ET AL.
by guest on January 21, 2016http://toxsci.oxfordjournals.org/
Dow
nloaded from