54, 262–273 (2000)
Copyright © 2000 by the Society of Toxicology
TOXICOLOGICAL SCIENCES
Acute Hepatotoxicant Exposure Induces TNFR-Mediated Hepatic Injury
and Cytokine/Apoptotic Gene Expression
Thomas L. Horn,* ,† Timothy D. O’Brien,‡ Lawrence B. Schook,* ,† and Mark S. Rutherford* ,† ,1
*Toxicology Graduate Program, †Department of Veterinary PathoBiology and ‡Department of Veterinary Diagnostic Medicine,
University of Minnesota, St. Paul, Minnesota 55108
Received August 30, 1999; accepted November 8, 1999
Tumor necrosis factor receptor knockout (TNFR KO) mice were
used to examine the role of tumor necrosis factor-␣ (TNF␣) signaling
during acute hepatotoxicant exposure. Mice were exposed intraperitoneally (ip) to either vehicle, phosphate-buffered saline (PBS), or
dimethylnitrosamine (DMN, 100 mg/kg) for 24 h. Histological evaluation showed that DMN-treated TNFR-2 KO mice had increased
liver damage compared to wild type (WT), TNFR-1 KO, or TNFR
double KO (DKO) mice. Also, 3 of 8 TNFR-2 KO mice died following
DMN treatment, suggesting that hepatic TNFR-2 signaling produces
protective responses that counteract TNFR-1-mediated damage.
DMN-induced cellular infiltration was absent in TNFR-1-deficient
mice, indicating that infiltrating cells do not exacerbate acute hepatotoxic events. In separate experiments, mice were exposed ip to
either DMN (5.0 or 100 mg/kg), carbon tetrachloride (CCl 4, 0.3 or 1.0
ml/kg), or corresponding PBS/corn oil controls for 6 or 24 h to
compare the hepatic mRNA expression of cytokine- and apoptoticassociated genes. Following 24 h of DMN (100 mg/kg) or 6 –24 h of
CCl 4 treatment, hepatic transcripts for TNF␣, interferon (IFN)-␥, IL
(interleukin)-1RI, and transforming growth factor (TGF)-␤RII were
induced. Hepatotoxicant-treated WT and TNFR DKO mice induced
liver transcripts for the pro- and anti-apoptotic genes, Bax and BclX L, respectively, indicating TNF-independent gene activation. The
anti-apoptotic gene, Bfl-1, was highly expressed in CCl 4-treated,
TNFR-positive strains, but minimally expressed in TNFR DKO
mice, suggesting that hepatic Bfl-1 is TNF-regulated. Taken together,
these data show that acute hepatotoxicant exposure is followed by
upregulation of liver cytokine, cytokine receptor, and apoptotic transcripts, and that TNF␣ regulates various aspects of liver inflammation and injury in a TNFR-specific fashion.
Key Words: apoptosis; Bfl-1; carbon tetrachloride (CCl 4); cytokines; dimethylnitrosamine (DMN); hepatotoxicity; inflammation;
necrosis; TNF.
Tumor necrosis factor (TNF) ␣ is a pleiotropic cytokine
critically involved in inflammation and immunity. TNF␣, in
Portions of this work were presented at the 1997 and 1999 annual meetings
of the Society of Toxicology.
1
To whom correspondence should be addressed at Department of Veterinary PathoBiology, University of Minnesota, 295e Animal Science/Veterinary
Medicine, 1988 Fitch Avenue, St. Paul, MN 55108. Fax: (612) 625– 0204.
E-mail: [email protected].
conjunction with interleukin (IL)-6, regulates the acute-phase
response, adhesion molecule activation, and antioxidant gene
expression. TNF␣ is perhaps the most critical and powerful
mediator of inflammation, cellular injury, cell death/apoptosis,
and wound healing (reviewed in Edwards et al., 1994).
There appears to be a critical role for TNF␣ in various
models of hepatotoxicity. For example, carbon tetrachloride
(CCl 4)-induced hepatotoxicity is blocked by the administration
of soluble TNF␣ receptors (TNFRs) (Czaja et al., 1995). Liver
repair following CCl 4 is mediated by TNF␣ (Bruccoleri et al.,
1997; Yamada and Fausto, 1998). In cadmium-induced hepatotoxicity, pretreatment with anti-TNF␣ antibodies prevented
focal inflammation as well as secretion of the acute-phase
reactant, serum amyloid A (Kayama et al., 1995), indicating
that these processes are cytokine-dependent. Similarly, neutralizing antibodies to TNF␣ delayed increases in serum levels of
IL-1␣ and liver enzymes as well as shortened the recovery time
following acetaminophen treatment (Blazka et al. 1995; 1996).
Acute inflammatory responses to 2,3,7,8-tetrachlorodibenzo-p-dioxin were mimicked by the administration of
exogenous IL-1␤ and TNF␣ (Moos et al., 1994). Taken together, these data suggest that TNF␣ contributes to pathological manifestations of chemical-induced liver damage. However, the mechanism by which TNF␣ controls hepatotoxicity
remains undefined.
There are 2 cell surface receptors for TNF␣ of approximately 55 kDa (TNFR-1) and 75 kDa (TNFR-2), each with
distinct functions (Tartaglia and Goeddel, 1992). TNFR-1 and
TNFR-2 lack significant homology in the intracellular domains, further suggesting that they are involved in differential
signaling pathways. For example, TNFR-1 is responsible for
the upregulation of c-fos, IL-6 and manganese superoxide
dismutase mRNA, prostaglandin E 2 synthesis, surface expression of IL-2 receptor and MHC class I and II molecules,
growth inhibition and cell death (Brakebusch et al., 1992;
Engelman et al., 1990; Espevik et al., 1990; Hohmann et al.,
1990; Kruppa et al., 1992; Naume et al., 1991; Shalaby et al.,
1990; Tartaglia et al., 1991; Thoma et al., 1990). TNFR-2
signaling seems to be restricted to T-cell proliferation (Grell et
al., 1998; Tartaglia et al., 1991); however, recent evidence
262
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
suggests that TNFR-2 mediates migration of Langerhans’ cells
(Wang et al., 1996; 1997).
Dimethylnitrosamine (N-nitrosodimethylamine, DMN) is a
member of the family of N-nitroso compounds that has potent
hepatotoxicant activity in mammals (Craddock, 1983) and is a
suspected carcinogen in humans. Lifetime exposure to nitrosamines occurs via dietary sources (cheeses, and cured meats
and fish), alcoholic beverages, and cigarette smoke, as well as
in industrial situations. The primary site for DMN metabolism
and toxicity is the liver. DMN bioactivation is thought to occur
either through the liver mixed-function oxidases (cytochrome
P450 2E1) or through hepatic monoamine oxidase (reviewed in
Myers and Schook, 1996). The end result is the formation of
toxic intermediates such as hydroxyl radicals, reactive oxygen
intermediates, and formaldehyde, with the consequence being
lipid peroxidation and/or DNA alkylation (Montesano and
Bartsch, 1976). Hepatic centrilobular necrosis has been observed in both mice and humans exposed to DMN.
Molecular mechanisms accounting for DMN-induced hepatotoxicity have been studied. Previously, we have demonstrated that exposure to DMN in vivo induces an acute-phase
response for serum amyloids A and P and albumin (Lockwood
et al., 1994). DMN exposure also results in a transient increase
in serum cytokine levels for IL-1␣, TNF␣, and IL-6 (Schook et
al., 1992). These data show that DMN exposure is associated
with hepatic proinflammatory gene expression. However, previous studies may have suffered from incomplete neutralization of TNF␣ activity, especially membrane-associated TNF␣.
Further, there has been no distinction between activities mediated by TNFR-1 versus TNFR-2.
To determine the signaling mechanism for TNF␣ during
acute liver injury, strains of TNFR knockout (KO) mice, lacking either TNFR-1, TNFR-2, or both receptors were exposed to
vehicle (phosphate-buffered saline [PBS]) or DMN (100 mg/
kg) for 24 h. Hepatic histopathological and cellular changes
during DMN exposure were examined using TNFR KO mice
compared against each TNFR KO strain as well as to TNFR
intact wild type (WT) mice. In addition, multi-probe RNase
protection assays (RPAs) were used to examine the hepatic
cytokine, cytokine receptor and apoptotic gene expression patterns following low- and high-dose acute DMN exposure in
WT and TNFR KO mice. Because both DMN and CCl 4 are
bioactivated to reactive intermediates by cytochrome P450 2E1
(Guengerich, 1987, 1991; Haggerty and Holsapple, 1990;
Koop, 1992) and both cause hepatic centrilobular necrosis
(Butler and Hard, 1971; Rechnagel and Glende, 1973; Reynolds and Ree, 1971), WT and TNFR KO mice were injected
with either DMN or CCl 4 in order to further investigate the
molecular events associated with TNF␣ during different xenobiotic exposure regimens. Results suggest that, overall, similar
mechanisms exist and TNF␣ shares a common role during
different xenobiotic exposure regimens.
263
MATERIALS AND METHODS
Mice and dosing regimens. Male and female WT (C57BL/6) mice (Harlan
Sprague Dawley, Indianapolis, IN) and male and female TNFR KO mice (a
gift from Immunex Corp., Seattle, WA) were housed in cages on a 12-h
light/dark cycle. TNFR-1 KO (p55 –/–p75 ⫹/⫹) mice have a C57BL/6 genetic
background while TNFR-2 KO (p55 ⫹/⫹p75 –/–) and TNFR double KO [DKO
(p55 –/–p75 –/–)] mice have a C57BL/6 ⫻ 129 background (Peschon et al., 1998).
DMN (Sigma Chemical Co., St. Louis, MO) was diluted from a 10 mg/kg
stock solution in PBS (pH 7) and stored at 4°C for less than 3 weeks. Solutions
were tested for endotoxin by the limulus amebocyte assay (Associates of Cape
Cod, Woods Hole, MA) prior to use. All mice (⬃20 g) were between 6 – 8
weeks old and were rested 1 week prior to dosing. Mice received 0.2 ml of
either vehicle (PBS) or DMN (at a final dose of 5.0 or 100 mg/kg) given as a
single daily injection intraperitoneally (ip) between 8 –10 am. CCl 4 (Sigma)
was diluted 0.3:10 or 1:10 in corn oil and mice were ip injected with 0.2 ml of
either corn oil or CCl 4 (0.3 or 1.0 ml/kg, respectively). All mice were sacrificed
6 or 24 h post-injection and livers removed. Endotoxin [lipopolysaccharide
(LPS), 50 ␮g E. coli 055:B55, Difco Laboratories, Detroit, MI] diluted in PBS
was administered ip for 6 h. All animal procedures followed standard protocols
approved by the University of Minnesota Animal Welfare Committee.
Histopathology. A portion of the liver was fixed in 4% buffered paraformaldehyde for 18 –24 h at 4°C and paraffin embedded. Serial sections (4 ␮m)
were mounted on poly-L-lysine-coated slides, deparaffinized and stained with
hematoxylin and eosin for separate analyses of histopathology and cellular
infiltration (below).
Histopathological and cellular infiltration analyses. Blinded liver samples were submitted for histological evaluation and at least 2 sections from
each liver were examined. Histopathological evaluation for necrosis included,
but was not limited to, assessing the distribution and extent of cell death
(hepatocytes and non-parenchymal cells). Lesions were graded on a semiquantitative scale: 0, normal; 1, rare foci of necrotic cells in centrilobular zones (no
more than 1–2 sites per section); 2, few necrotic foci (less than half of
centrilobular zones had sites of necrosis); 3, many/diffuse centrilobular zones
with necrosis (confined to centrilobular zone only); 4, diffuse centrilobular to
midzonal necrosis; and 5, diffuse submassive to massive necrosis (most or
almost all of lobule was necrotic).
The extent of inflammatory cellular infiltration into the liver was examined
using the same liver samples as for histopathology. Cellular infiltration was
graded visually on a semiquantitative scale: 0, no influx; 1, rare inflammatory
cells (no more than 1–3 cells in 1 or 2 centrilobular zones per section); 2, few
inflammatory cells (1–5 cells in less than 50% of centrilobular zones); 3,
moderate inflammatory cell infiltration (5–15 cells in most centrilobular
zones); 4, marked inflammatory cell infiltration (greater than 15 cells in most
centrilobular to midzonal areas); and 5, severe inflammatory cell infiltration
(too numerous to count and infiltrating most of lobule).
RNA extraction. Liver sections (⬃200 mg) were quickly thawed and
homogenized in 2 ml of TRIZOL (Life Technologies, Grand Island, NY) and
total RNA was isolated according to the manufacturer’s directions. RNA was
resuspended in RNase-free water, quantitated using UV spectrophotometry,
and stored at – 80°C.
Multi-probe RPAs. The DNA templates (mCK3, mCR4, mAPO2, and
mAPO3), and In Vitro Transcription and RPA kits were purchased from
PharMingen (San Diego, CA). Briefly, RNA probes were generated using 1 ␮g
of template DNA and in vitro transcribed using T7 RNA polymerase. Yeast
tRNA (10 ␮g) and PharMingen-supplied mouse control RNA (2 ␮g) were used
as negative and positive controls, respectively. Total liver RNA (10 –20 ␮g)
was dried and resuspended in 2.4 – 4 ⫻ 105 cpm/␮g and hybridized at 56°C for
16 –18 h. Following hybridization, RNA samples were RNase A/T1-digested
for 45 min at 30°C, followed by incubation with proteinase K for 15 min at
37°C and then ethanol precipitated. Protected RNAs were separated on a 6%
polyacrylamide/7 M urea denaturing gel. Following electrophoresis, the gels
were dried and exposed to X-ray film or phosphorimager screens and bands
264
HORN ET AL.
were quantitated using densitometry or phosphorimagery, respectively (Molecular Dynamics, Sunnyvale, CA).
Statistical analyses. All necrosis and cellular infiltration experiments were
performed with at least 5 mice and the results are represented as an average ⫾
SE for each mouse strain within each dosing group. The results were analyzed
using the Kruskal-Wallis H test, a nonparametric equivalent to a 1-way
analysis of variance. A 95% confidence level was used to define statistical
significance. The hepatic gene expression data shown is from one group of
mice and results are representative for 2 independent, identically treated groups
(n ⫽ 2).
RESULTS
Histological Analyses Following Acute Exposure to DMN
Previous studies using various hepatotoxicants such as cadmium (Kayama et al., 1995) and acetaminophen (Blazka et al.,
1995; 1996) have focused on exposure to large acute doses in
which TNF␣ has been shown to play a role in hepatotoxicity.
Therefore, WT and TNFR KO mice received either an acute
vehicle (PBS) or DMN (100 mg/kg) dose ip and were sacrificed 24 h post-exposure. This DMN dose has been previously
shown to produce both liver necrosis and hepatic DNA fragmentation (Ray et al., 1992).
All vehicle-treated mice had normal liver histology irrespective of TNFR genotype (Fig. 1). Centrilobular and midzonal
necrosis was evident in all WT and TNFR KO mice exposed to
DMN. Hepatic cords in affected areas were often disrupted and
exhibited individualization of hepatocytes. Hepatocytes in affected centrilobular zones were characterized by clumped chromatin, karyorrexis, or nuclear pyknosis, and by hypereosinophilic cytoplasm. In more severely affected areas, there was
loss of hepatocytes with centrilobular to midzonal hemorrhage
(Fig. 1).
Necrosis and Cellular Infiltration following DMN Exposure
To further characterize the observed changes following an
acute dose of DMN (100 mg/kg), liver lesions for each treatment group were evaluated using a semiquantitative scale, as
described in the Materials and Methods. However, since we
observed that WT mice exposed to DMN contained inflammatory cell foci, which was reduced up to 50% in the TNFR KO
strains (data not shown), it was necessary to develop a scoring
procedure to separately analyze the degree of liver necrosis
independently from hepatic cellular infiltration.
All vehicle-treated WT and TNFR KO mice had necrosis
and cellular infiltration scores of 0 (data not shown). Necrosis
and cellular infiltration scores for mice exposed to DMN (100
mg/kg) are shown in Figure 2. No significant differences in
necrosis were observed between WT, TNFR-1 KO, and TNFR
DKO mice exposed to DMN. However, TNFR-2 KO mice had
significantly (p ⬍ 0.05) higher necrosis scores (3.4 ⫾ 0.2)
compared to WT (2.6 ⫾ 0.2), TNFR-1 KO (2.4 ⫾ 0.2) and
TNFR DKO (2.9 ⫾ 0.1) mice (Fig. 2A). In addition, 37.5%
(3/8) of TNFR-2 KO mice succumbed to DMN-induced death
prior to tissue harvest at 24 h post-exposure (dead TNFR-2 KO
mice were not included in the scoring procedure). We also
observed an almost complete absence of inflammatory cells in
TNFR-1 KO and TNFR DKO mice (Fig. 2B), indicating that
TNFR-1 mediates DMN-induced cellular influx into the liver.
These infiltrating cells were observed to be mainly neutrophils,
macrophages, and T cells (data not shown).
Hepatotoxicant-Induced Cytokine and Cytokine Receptor
mRNA Expression in WT Mice
Because TNFR signals appear to differentially regulate inflammatory-cell influx during DMN exposure, we determined
hepatic proinflammatory cytokine gene expression in the same
livers as above. Multi-probe RPA mCK3 and mCR4 template
DNAs were used to quantitate hepatic cytokine and cytokine
receptor mRNA expression, respectively following acute hepatotoxicant exposure. In addition, to compare hepatic gene expression for different hepatotoxicants, mice were acutely exposed to either DMN (5.0 or 100 mg/kg) or CCl 4 (0.3 or 1.0
ml/kg) for up to 24 h. For comparison, LPS (50 ␮g) was
administered to stimulate hepatic cytokine/cytokine receptor
gene expression.
The proinflammatory cytokines TNF␣, lymphotoxin (LT)␤,
TNF␣, IL-6, interferon (IFN)-␥, IFN-␤, transforming growth
factor (TGF)-␤1 and TGF␤2 were analyzed by RPA (Fig. 3).
As expected, LPS increased hepatic transcripts for TNF␣ and
IL-6. At 6-h post-exposure, vehicle-treated animals showed
enhanced transcripts for TNF␤ and TNF␣ compared to naive
and DMN (5.0 mg/kg)-treated mice, and increased IFN-␥ transcripts as compared to DMN-treated mice. Increased cytokine
expression observed following 6 h of vehicle treatment is
consistent with a transient increase in inflammatory gene expression that subsides by 24 h (Lockwood et al., 1991; 1994;
Schook et al., 1992). Following 24 h of high-dose DMN (100
mg/kg), hepatic expression of TNF␤, LT␤, TNF␣, and IFN-␥
was induced 3- to 10-fold compared to vehicle controls, with
little change in IL-6, IFN-␤, TGF␤1, or TGF␤2 (Fig. 3).
However, compared to naive mice, only transcripts for TNF␣
were greatly increased following DMN (100 mg/kg).
In contrast to what we observed for 6-h DMN (5.0 mg/kg),
hepatic transcripts for TNF␤, TNF␣, IFN-␥, and TGF␤1 were
higher following 6-h CCl 4 (0.3 or 1.0 ml/kg) as compared to
corn oil controls, whereas LT␤ was induced only following 6-h
CCl 4 (1.0 ml/kg) (Fig. 4). Hepatic cytokine expression following CCl 4 exposure showed similar results as compared to 24-h
DMN (100 mg/kg). Following 24-h CCl 4 (0.3 or 1.0 ml/kg),
hepatic cytokine mRNA expression for TNF␣, IFN-␥, LT␤,
and TGF␤1 was still upregulated compared to corn oil controls. In addition, however, we also observed that hepatic
TGF␤2 transcripts were induced following 24-h CCl 4 (0.3 or
1.0 ml/kg) compared to 6-h CCl 4 treatment and corn oil controls. In contrast to DMN exposure, CCl 4-treated mice had
upregulated TGF␤1 and TGF␤2 transcript levels, demonstrat-
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
ing that these hepatotoxicants differentially regulate inflammation.
Importantly, we observed differences in hepatic cytokine
expression following PBS and corn oil treatments (Figs. 3 and
4). Whereas PBS induced a transient increase in hepatic transcripts for TNF␣ and TNF␤ (compared to naive), corn oil
exposure had little effect on these cytokines. Compared to
naive mice, 24-h corn oil treatment suppressed transcripts for
TNF␤, LT␤, and TGF␤1, similar to what was observed for
24-h PBS exposure. These data further demonstrate that vehicle effects must be considered during interpretation of molecular data and that both naive and vehicle controls must be
considered.
Because cytokines must signal through specific receptors in
order to elicit biological effects, proinflammatory cytokine
receptor transcript levels were analyzed (Fig. 5). Differences in
hepatic cytokine receptor expression in WT mice were not
observed between vehicle and DMN (5.0 mg/kg) following a
6-h exposure. In contrast, a high-dose 24-h DMN (100-mg/kg)
exposure induced hepatic expression of IL-1RI and IL-1RII,
but suppressed gp130 transcripts compared to both naive and
vehicle controls (Fig. 5A). IL-6R␣ transcripts were not significantly altered during any vehicle or DMN exposure. In contrast to DMN, CCl 4 (0.3 or 1.0 ml/kg) was a potent inducer of
hepatic cytokine receptor expression in WT mice. Mice treated
with CCl 4 (0.3 or 1.0 ml/kg) showed upregulated transcripts for
IL-1RI, TNFR-2, TNFR-1, IL-6R␣, gp130, TGF␤RI and
TGF␤RII at both 6 and 24 h. Interestingly, and in contrast to
DMN, CCl 4 did not alter expression of IL-1RII (Fig. 5B).
These data further confirm differential effects of cytokine
production and signaling in the liver for these 2 hepatotoxicants.
Apoptotic Liver mRNA Expression following Hepatotoxicant
Exposure in WT and TNFR KO Mice
High-dose DMN (100-mg/kg) exposure initiates apoptosis
in the liver (Ray et al., 1992). A link between TNF␣ and
apoptosis within the liver has been established (Leist et al.,
1995; 1997; Rolfe et al., 1997; Takehara et al., 1998), and
apoptosis is one proposed contributor to xenobiotic-induced
liver injury. Thus, the regulation of apoptotic promoting genes,
Fas, Bcl-X S, Bak, Bax, and Bad, and anti-apoptotic genes,
Bcl-W, Bfl-1, Bcl-X L,, and Bcl-2 were measured by RPA. To
abrogate TNF␣ signals, WT and TNFR DKO mice were compared following acute exposure to either DMN (5.0 or 100
mg/kg) or CCl 4 (0.3 or 1.0 ml/kg) for 6 –24 h.
Compared to vehicle and naive controls, exposure of WT
mice to DMN (5.0 mg/kg) for 6 h did not alter hepatic Fas
transcripts. In contrast, TNFR DKO mice exposed to DMN
(5.0 mg/kg) exhibited a 2.5-fold increase in Fas expression
compared to vehicle controls (Fig. 6A). Exposure to high-dose
DMN (100 mg/kg) for 24 h increased (1.5- to 1.7-fold) hepatic
Fas expression in both WT and TNFR DKO mice compared to
vehicle controls.
265
Because we observed differences in inflammatory cytokine
and receptor gene expression for DMN and CCl 4 treatment, we
examined whether these would impact apoptotic gene expression. Exposure of WT mice to CCl 4 (0.3 and 1.0 ml/kg) did not
induce hepatic Fas expression (Fig. 6B). Hepatic Fas was
decreased (1.9- to 3.5-fold) in TNFR DKO mice exposed to
CCl 4 at both 6 and 24 h (Fig. 6B). The fact that Fas was
expressed in the liver in both WT and TNFR DKO mice
suggests that TNF␣ is not required for hepatic Fas expression,
but may regulate Fas mRNA levels.
Hepatic expression of the Bcl-2 family of genes was measured in WT and TNFR DKO mice exposed to DMN (5.0 or
100 mg/kg) for 6 or 24 h, respectively (Table 1). We observed
that naive TNFR DKO mice had 41% fewer relative Bcl-X L
transcripts. Hepatic Bcl-X L and Bax expression were increased
following low- and high-dose DMN in both WT and TNFR
DKO mice, whereas a down-regulated expression (37– 68%) of
Bcl-W was observed at 24 h in both strains of mice. Only
minor differences between vehicle and DMN in WT or TNFR
DKO were observed for Bfl-1 and Bak (Table 1). Hepatic
Bcl-2 was not detected following vehicle or DMN exposure in
either strain (data not shown).
Hepatic expression of the Bcl-2 family of genes was also
measured in WT and TNFR DKO mice exposed to CCl 4 (Table
2). Similar to DMN exposure, hepatic mRNA expression of
Bcl-X L and Bax was induced for all CCl 4 exposures in both
WT and TNFR DKO mice compared to corn oil controls. In
contrast to DMN, WT and TNFR DKO mice exposed to CCl 4
(0.3 or 1.0 ml/kg) showed increased (1.7- to 3.8-fold) liver
expression of Bcl-W at 24 h compared to corn oil controls.
Also, whereas DMN exposure did not alter hepatic Bfl-1 transcripts in either mouse strain, we observed significant levels of
Bfl-1 transcripts following 24 h CCl 4 exposure in WT, but not
TNFR DKO mice (Table 2). This again demonstrates intricate
differences for these two hepatotoxicants, and suggests that
hepatic Bfl-1 expression may be TNF-mediated in certain
conditions.
To further characterize how TNF␣ regulates hepatic Bfl-1
gene expression, WT, TNFR-1 KO, TNFR-2 KO and TNFR
DKO mice were exposed to corn oil or CCl 4 (0.3 or 1.0 ml/kg)
for 6 or 24 h and hepatic Bfl-1 transcripts were then quantitated
by RPA. Data showed that CCl 4 exposure upregulates liver
Bfl-1 in WT, TNFR-1 KO and TNFR-2 KO mice, with little
alteration in TNFR DKO mice (Fig. 7). This suggests that
hepatic Bfl-1 is TNF-mediated during CCl 4 exposure, but that
either TNFR can mediate its upregulation during acute CCl 4
exposure.
DISCUSSION
The inflammatory response associated with chemical-induced hepatotoxicity, including DMN, is characterized by hepatic proinflammatory cytokine and acute-phase reactant gene
expression (Bhattacharjee et al., 1998; Horn et al., 1998;
266
HORN ET AL.
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
FIG. 2. Necrosis and cellular infiltration scores following acute DMN
exposure. WT and TNFR KO mice received an acute DMN (100-mg/kg)
exposure ip and were sacrificed 24 h post-exposure. Livers were removed,
processed for paraffin embedding, and liver sections graded using a semiquantitative scale as described in Materials and Methods. The individual necrosis
(A) and cellular infiltration (B) scores from each mouse strain were determined
(average ⫾ SE). 0 ⫽ score of 0 for that mouse strain. All vehicle-treated
animals had scores of 0 (data not shown). A minimum of 5 animals per group
was used. (Significant difference [p ⬍ 0.05]: ⫹, compared to WT; triangle,
compared to TNFR-1 KO; #, compared to TNFR DKO).
Lockwood et al., 1994; Schook et al., 1992). Aberrant TNF␣
production is associated with DMN exposure and is consistent
with reports demonstrating a role for TNF␣ in DMN hepatotoxicity (Schook et al., 1992). These observations have led to
the hypothesis that toxic intermediates, produced both as a
result of xenobiotic metabolism and the inflammatory response, contribute to the hepatic centrilobular necrosis that
occurs following toxicant exposure. While inflammation and
hepatic TNF␣ are thought to be part of the healing process
(Akerman et al., 1992; Bruccoleri et al., 1997; Kimura et al.
1997), an uncontrolled inflammatory response may be detrimental to the host. Accumulation of liver immunocytes (neutrophils, monocytes/macrophages, and/or T cells) may lead to
additional hepatocyte death through production and release of
reactive oxygen intermediates that can damage liver architecture. Conversely, increased immunocyte influx may also be
needed to clear dead or dying hepatocytes in order for liver
regeneration to occur.
In order to understand the cellular and molecular effects of
267
TNF␣ in the liver during acute chemical-induced hepatotoxicity, WT, TNFR-1 KO, TNFR-2 KO and TNFR DKO mice
were exposed to high-dose DMN (100 mg/kg). Moreover,
acute exposure to another hepatotoxicant, CCl 4, was compared
at the level of hepatic gene expression. In order to focus on
apoptotic genes with minimal necrosis, a dose of 0.3 ml/kg
CCl 4 was given for 6 h as previously discussed (Shi et al.,
1998). Similarly, a low-dose exposure to DMN (5.0 mg/kg)
(Bhattacharjee et al., 1998; Lockwood et al., 1994; Myers and
Schook, 1996) was used because necrosis is not detected
following 6 h of treatment (manuscript in preparation). A 24 h
exposure to CCl 4 (1.0 ml/kg) was used to induce hepatic
necrosis (Yamada and Fausto, 1998) with levels comparable to
our high-dose DMN (100 mg/kg) exposure regimen (T. D.
O’Brien, unpublished results). Hence, we were able to compare
both low- and high-dose acute exposures for these xenobiotics
to reveal subtle differences in hepatic gene regulation and the
role of TNF␣.
Previous reports have used antibodies or soluble receptors
directed against TNF␣ in order to neutralize its effect (Blazka
et al., 1995; 1996; Bruccoleri et al., 1997; Czaja et al., 1995;
Kayama et al., 1995), but this approach can be problematic
because the neutralization may not be 100%, particularly during chronic exposure regimens. Further, the reagents may not
neutralize membrane bound TNF␣. In addition, because each
TNF␣ receptor has structurally dissimilar intracellular domains
(Tartaglia and Goeddel, 1992), differential receptor signaling
may regulate separate hepatotoxic or repair events. Indeed,
TNFR-1 appears to be required for the rapid influx of immunocytes into the liver during high-dose, acute DMN exposure.
We have used this model to show that immunocyte influx was
not required for liver pathology to occur during acute DMN
exposures. Our results further suggested that TNFR-1 signals
may be detrimental in the liver unless TNFR-2 signals also
occur (Fig. 2). Thus, genetic TNFR KO mice permit the study
of TNF␣ effects with respect to individual receptor signaling
events.
Acute exposure to DMN or CCl 4 induced a proinflammatory
response accompanied by induction of hepatic TNF␣ and
IFN-␥ (Figs. 4 and 5). Thus, the liver is a potential source for
TNF␣ detected in the serum following low-dose, multiple
exposures to DMN (Schook et al., 1992). In addition, CCl 4
rapidly upregulated hepatic expression of TNFR-1 and
TNFR-2, whereas the effect of DMN on these transcripts was
indistinguishable from vehicle controls (Fig. 5). Our data further substantiates that TNF␣ is a key cytokine involved in
FIG. 1. Hepatic histopathology following acute DMN exposure in vivo. WT and TNFR KO mice received an acute ip injection of either vehicle (PBS) or
DMN (100 mg/kg) and were sacrificed 24 h post-injection. Livers were removed and processed for hematoxylin and eosin staining as described in Materials and
Methods. Liver photomicrographs were from one group of mice exposed to either vehicle or DMN and are representative of 5 mice for each exposure group.
Note disruption of hepatic cord architecture along with hepatocellular necrosis (black arrows) in all DMN-treated mice. Prominent hemorrhage was also visible
in the sections from WT, TNFR-1 KO and TNFR-2 KO mice. Inflammatory cells are apparent in the DMN-treated WT and TNFR-2 KO liver sections (white
arrows). Original magnification ⫻ 82.5; hepatic central vein, C.
268
HORN ET AL.
FIG. 3. Hepatic proinflammatory cytokine expression following acute
DMN exposure. WT mice were exposed to either vehicle (PBS) or DMN for
6 or 24 h. Liver RNA (20 ␮g) was isolated and used for RPA with mCK3
template DNA (PharMingen). Results shown are representative for 2 independent identically-treated groups. (A) All RNA samples were run on the same
gel. However, the gel was exposed to X-ray film for 3 days for high-copy
mRNAs (L32, GAPDH, diluted probes and positive control) or 10 days to
detect low-copy transcripts (hepatic cytokines). DP, diluted undigested RPA
probes (ladders); Y, yeast tRNA (negative control); P, PharMingen control
RNA (2 ␮g); N, naive; L, LPS (50 ␮g dose for 6 h); V, vehicle; D, DMN
(subscript numbers represent the dose in mg/kg). (B) Relative cytokine mRNA
abundance was determined using densitometry. Results were normalized to the
housekeeping gene, GAPDH, and were determined by dividing the cytokine
densitometry value by the GAPDH densitometry value and multiplying by 100.
FIG. 4. Hepatic proinflammatory cytokine expression following acute
CCl 4 exposure. WT mice were exposed to either corn oil or CCl 4 for 6 or 24 h.
Liver RNA (20 ␮g) was isolated and used for RPA with mCK3 template DNA
(PharMingen). Results shown are representative for 2 independent, identically
treated groups. (A) All RNA samples were run on the same gel. However, the
gel was exposed to X-ray film for 1 day for high-copy mRNAs (L32, GAPDH,
diluted probes and positive control) or 10 days to detect low-copy transcripts
(hepatic cytokines). Abbreviations: DP, diluted undigested RPA probes (ladders); Y, yeast tRNA (negative control); P, PharMingen control RNA (2 ␮g);
N, naive; L, LPS (50-␮g dose for 6 h); O, corn oil; C, CCl 4 (subscript numbers
represent the dose in ml/kg). (B) Relative cytokine mRNA abundance was
determined using phosphorimagery. Results were normalized to the housekeeping gene, GAPDH, and were determined by dividing the cytokine phosphorimagery value by the GAPDH phosphorimagery value and multiplying by
100.
DMN and CCl 4 hepatotoxicity. However, our results also
showed that there are subtle, yet critical differences in the
hepatic gene expression for these diverse xenobiotics despite a
common bioprocessing pathway. This may result from the
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
269
TNF␣ dependence for DMN-mediated hepatotoxicity in acute
high-dose DMN (100 mg/kg) versus subacute low-dose DMN
(5.0 mg/kg) exposure regimens (manuscript in preparation).
Finally, whereas genetic knockouts completely abrogate TNF␣
signals, it cannot be excluded that soluble TNF␣ receptors did
not completely neutralize TNF␣ signals; small amounts of
TNF␣ may signal protective responses that are masked in the
presence of elevated TNF␣ signaling. Although we cannot as
yet link Bfl-1 expression to hepatotoxicity, it is interesting that
CCl 4 differed from DMN by upregulating TNFR-1, TNFR-2
and Bfl-1 transcripts, the latter in a TNFR-dependent manner.
Thus, our data show differences exist between hepatic cyto-
FIG. 5. Hepatic proinflammatory cytokine receptor expression following
acute hepatotoxicant exposure. WT mice were exposed to either (A) vehicle
(PBS) or DMN or (B) corn oil or CCl 4 for 6 or 24 h. Liver RNA (10 ␮g) was
isolated and used for RPA with mCR4 template DNA (PharMingen). Results
are representative of 1 of 2 independent groups. Relative cytokine mRNA
abundance was determined using (A) densitometry or (B) phosphorimagery.
Results were normalized to the housekeeping gene, GAPDH, and were determined by dividing the cytokine receptor value by the GAPDH value and
multiplying by 100.
different bioreactive intermediates that are produced from each
compound. Our histological data show that TNFR-1-deficient
mice exposed to high-dose DMN lack hepatic cellular infiltration, yet show no reduction in necrosis compared to DMNtreated WT mice (Fig. 2). However, a previous report observed
decreases in both CCl 4-mediated hepatotoxicity and inflammation following pretreatment with soluble TNFR (Czaja et al.,
1995). These differences could be related to the high dose (5.0
ml/kg) of CCl 4 used in that study, as was previously suggested
(Bruccoleri et al., 1997; Louis et al., 1998). This high dose of
CCl 4 could result in extensive hepatic necrosis and TNF␣
production. In addition, important differences in the metabolites generated from CCl 4 and DMN may also influence subsequent TNF␣ effects. We have also observed differences in
FIG. 6. Hepatic Fas mRNA expression following acute hepatotoxicant
exposure. WT and TNFR DKO mice were exposed to either (A) vehicle (PBS)
or DMN (5.0 or 100 mg/kg) or (B) corn oil or CCl 4 (0.3 or 1.0 ml/kg) for 6 or
24 h. Liver RNA (10 ␮g) was isolated and used for RPA with mAPO3 template
DNA (only Fas is shown). Hepatic Fas expression was normalized to GAPDH
by dividing the Fas densitometry/phosphorimagery value by the GAPDH
densitometry/phosphorimagery value and multiplying by 100. Representative
results from one of two experiments are shown. Abbreviations: N, naive; L,
LPS (50 ␮g dose for 6 h); V, vehicle; D 5 or D 100, DMN (5.0 or 100 mg/kg,
respectively); O, corn oil; C 0.3 or C 1.0 (CCl 4, 0.3 or 1.0 ml/kg, respectively).
270
HORN ET AL.
TABLE 1
Hepatic Apoptotic mRNA Expression following Acute Dimethylnitrosamine Exposure
6h
24 h
Apoptotic Gene a
Naive
LPS
PBS b
DMN (5.0)
PBS
DMN (100)
WT Mice
Bcl-W
Bfl-1
Bcl-X L
Bak
Bax
18.5 c
0.9
49.8
7.0
16.4
22.1
13.6
122.7
8.9
47.1
20.9
3.3
54.4
7.7
22.4
18.4
3.2
65.3
6.0
38.3
18.7
2.1
52.5
3.8
19.6
11.9
3.5
122.2
6.3
66.5
TNFR DKO mice
Bcl-W
Bfl-1
Bcl-X L
Bak
Bax
18.6
1.5
29.1
5.4
11.6
7.5
5.9
60.9
7.0
17.5
14.0
2.2
19.3
5.9
10.2
20.4
2.5
39.4
9.3
30.0
19.7
2.1
33.0
6.1
12.1
6.4
1.9
43.4
5.6
38.7
Note. WT and TNFR DKO mice were exposed to either vehicle (PBS) or DMN ip for 6 or 24 h. In addition, naive (no treatment) or 6-h LPS (50 ␮g)-exposed
mice were used as controls. Livers were removed and total liver RNA was isolated. RNA (10 ␮g) was then used with mAPO2 multi-probe RPA to detect hepatic
Bcl-2 family of apoptotic genes.
a
Bcl-W, Bfl-1 and Bcl-X L are anti-apoptotic, while Bak and Bax promote apoptosis.
b
Mice were exposed ip to either vehicle (PBS) or DMN (5.0 mg/kg) for 6 h or exposed to vehicle or DMN (100 mg/kg) for 24 h.
c
Results from a single experiment are shown and are representative of 2 different identically treated mice. Results were normalized to the housekeeping gene,
GAPDH, and were determined by dividing the apoptosis densitometry/phosphorimagery value by the GAPDH densitometry/phosphorimagery value and
multiplying by 100.
kine, cytokine receptor, and apoptotic genes following acute
DMN and CCl 4 exposures, and that the role of TNF␣ may be
more critical for mediating specific pathologic and molecular
outcomes in each system.
TNF␣ has been shown to down-regulate P450 activities by
decreasing their expression (Ghezzi et al., 1986; Renton and
Armstrong, 1994; Warren et al., 1999). Thus, the use of TNFR
DKO mice may be problematic because the lack of TNF␣
signaling could lead to uncontrolled P450-mediated bioactivation of DMN or CCl 4 leading to increased hepatotoxicity.
However, Boess et al. (1998) have shown that P450 2E1
activity, which is the principle enzyme responsible for the
conversion of acetaminophen to its hepatotoxic metabolite
(Lee et al., 1996; Thomsen et al., 1995), was not altered
following acute exposure to acetaminophen (400 mg/kg) in
TNF␣/LT␣ double-deficient mice. This would suggest that the
lack of TNF signaling does not alter the metabolic machinery
during DMN exposure. Our results support this hypothesis,
because TNFR DKO mice did not show increased hepatotoxicity following a 24-h DMN (100 mg/kg) exposure compared
to WT, TNFR-1 KO and TNFR-2 KO (Fig. 2A).
Following TNF␣ administration, vascular cell adhesion molecule-1 and E-selectin up-regulation does not occur in TNFR-1
KO mice, as compared to WT mice (Neumann et al., 1996).
TNF␣ challenge caused mononuclear and neutrophil infiltration in the liver, lungs, and kidneys in WT mice, but not in
TNFR-1 KO mice, suggesting that the TNFR-1 is responsible
for mediating leukocyte infiltration in vivo. In agreement with
this, during an acute DMN (100 mg/kg) exposure, cellular
infiltration was almost completely absent in TNFR-1 KO mice,
and was completely absent in TNFR DKO as compared to WT
and TNFR-2 KO mice (Fig. 2B). Importantly, our results also
show that high-dose DMN-induced hepatotoxicity can occur in
the absence of liver inflammation. TNFR-1 KO and TNFR
DKO mice had similar necrosis scores to WT mice despite the
paucity of inflammatory cell accumulation (Fig. 2).
Recently, Peschon et al. (1998) observed decreased pulmonary neutrophil influx in TNFR-1 KO and TNFR DKO mice
and exacerbated neutrophil accumulation in the lungs of
TNFR-2 KO mice following Micropolyspora faeni infection.
They concluded that TNFR-1 is responsible for neutrophil
influx during bacterial-induced pulmonary inflammation and
that TNFR-2 may be important for down-regulating the inflammatory response. These data further support our acute, highdose DMN experiments in which TNFR-1 KO and TNFR
DKO mice had little cellular infiltrates, while TNFR-2 KO
mice had levels comparable to WT (Fig. 2B). Hepatotoxicity
following an acute DMN (100-mg/kg) dose is significantly
higher in TNFR-2 KO mice compared to WT, TNFR-1 KO,
and TNFR DKO. Interestingly, TNFR-2 KO mice were apparently more susceptible to acute DMN (100 mg/kg)-induced
lethality, since 37.5% were found dead prior to sacrifice. Because these dead TNFR-2 KO mice were not included in the
scoring procedure, the level of necrosis presented in Figure 2A
may be underestimated. This further suggests that TNFR-2
signaling may be providing some protective responses because
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
271
TABLE 2
Hepatic Apoptotic mRNA Expression following Acute Carbon Tetrachloride Exposure
6h
24 h
Apoptotic gene a
Naive
LPS
Corn oil b
CCl 4 (0.3)
CCl 4 (1.0)
Corn oil
CCl 4 (0.3)
CCl 4 (1.0)
WT Mice
Bcl-W
Bfl-1
Bcl-X L
Bak
Bax
3.1 c
⬍0.1
7.7
1.1
5.5
3.1
1.4
7.1
3.3
8.4
2.5
ND d
6.3
1.5
4.3
2.4
0.6
22.6
1.4
7.9
2.7
0.7
26.3
1.4
7.5
2.5
⬍0.1
8.2
1.1
5.3
7.5
5.7
30.2
3.7
22.3
9.4
3.2
26.3
4.4
25.6
4.1
0.1
8.5
1.8
5.0
3.0
5.0
60.7
4.6
10.0
2.4
ND
8.0
1.0
4.7
4.2
0.2
26.2
2.7
12.2
2.4
0.3
31.1
1.7
10.1
2.2
⬍0.1
6.9
1.4
3.3
3.8
0.6
12.3
1.9
12.5
4.0
0.2
21.1
3.8
11.7
TNFR DKO mice
Bcl-W
Bfl-1
Bcl-X L
Bak
Bax
Note. WT and TNFR DKO mice were exposed to either corn oil or CCl 4 ip for 6 or 24 h. In addition, naive (no treatment) or 6 h LPS (50 ␮g)-exposed mice
were used as controls. Livers were removed and total liver RNA was isolated. RNA (10 ␮g) was then used with mAPO2 multi-probe RPA to detect hepatic Bcl-2
family of apoptotic genes.
a
Bcl-W, Bfl-1 and Bcl-X L are anti-apoptotic, while Bak and Bax promote apoptosis.
b
Mice were exposed ip to either corn oil or CCl 4 (0.3 or 1.0 ml/kg) for 6 or 24 h.
c
Results from a single experiment are shown and are representative of 2 different identically treated mice. Results were normalized to the housekeeping gene,
GAPDH, and were determined by dividing the apoptosis densitometry/phosphorimagery value by the GAPDH densitometry/phosphorimagery value and
multiplying by 100.
d
Not detected.
TNFR-1 KO mice had lower necrosis compared to TNFR-2
KO mice. Taken together, these data suggest that TNFR-1
delivers detrimental signals that may be counteracted by
TNFR-2-mediated signals. It has not been determined whether
FIG. 7. Hepatic expression of Bfl-1 following acute CCl 4 exposure. WT,
TNFR DKO, TNFR-1 KO and TNFR-2 KO mice were exposed to either corn
oil or CCl 4 (0.3 or 1.0 ml/kg) for up to 24 h. Liver RNA (10 ␮g) was isolated
and used for RPA with mAPO2 template DNA (only Bfl-1 is shown). Hepatic
Bfl-1 expression was normalized to GAPDH by dividing the Bfl-1 phosphorimagery value by the GAPDH phosphorimagery value and multiplying by 100.
Representative results from one of two independent experiments are shown.
Abbreviations: N, naive; L, LPS (50 ␮g dose for 6 h); V, vehicle (PBS); D 5 or
D 100, DMN (5.0 or 100 mg/kg, respectively); O, corn oil; C 0.3 or C 1.0 (CCl 4, 0.3
or 1.0 ml/kg, respectively); nd, not detected.
TNFR-2-deficient mice reflect a down-regulated inflammatory
response at cytokine transcript levels.
TNFR-deficient mice (same strains as used in our experiments) express 5-fold higher TNF␣ protein following LPS
exposure than do WT mice (Peschon et al., 1998), and this may
explain the elevated damage in TNFR-2 KO mice following
DMN (100 mg/kg) (Fig. 2A). Because TNFR-1 signaling is
still intact in TNFR-2 KO mice, increased hepatic expression
of TNF␣ following exposure to inflammatory mediators (LPS,
DMN or CCl 4) can act on TNFR-1 to initiate destructive
responses. However, it cannot be ruled out that the differences
we observed in necrosis scores may be due to signaling through
TNFR-2 that provides protective responses to counteract
TNFR-1-mediated damage.
Recent evidence indicates that high dose DMN exposure
(100 mg/kg) induces an apoptotic mechanism that may contribute to toxicity (Ray et al., 1992). However the hepatic
genes involved were not examined. We observed upregulation
of Bax and Bcl-X L following acute, high-dose DMN or CCl 4
exposure, supporting that regulation of these genes determines
the fate of apoptosis, independent of TNF signaling.
Bfl-1 (also called A1) has been shown to suppress p53induced apoptosis (D’Sa-Eipper et al., 1996). Previous reports
showed that Bfl-1 induction in cultured endothelial cells is
mediated by TNF␣, and increased expression of Bfl-1 protects
against TNF-mediated cell death (Karsan et al., 1996a; 1996b).
Consistent with this, hepatic Bfl-1 gene expression was up-
272
HORN ET AL.
regulated in a TNFR-dependent manner during CCl 4 but not
DMN exposure. High-dose DMN exposure slightly induced
hepatic Bfl-1 expression compared to naive but not vehicle
controls in WT mice (Table 1). In contrast, corn oil controls did
not upregulate Bfl-1 transcript levels, and Bfl-1 transcripts
were greatly elevated in CCl 4-treated livers from WT but not
TNFR DKO mice (Fig. 7 and Table 2). Thus, TNF␣ mediates
CCl 4-, but not DMN-induced hepatic expression of Bfl-1, and
either TNFR can provide the necessary signals.
In summary, through the utilization of TNFR KO mice, we
have shown that TNF␣ is a key mediator in hepatic immunocyte infiltration in addition to regulating necrotic events during
acute, high-dose chemical exposure. The outcome of hepatotoxicity and inflammation depends on specific TNFR signaling.
Further, both the molecular response and the role of TNF␣ in
regulating hepatic inflammatory cytokine, cytokine receptor,
and apoptotic gene expression is unique for DMN and CCl 4.
Thus, it may be necessary to individually evaluate the contribution of TNF␣ to individual hepatotoxic processes for various
xenobiotics.
ACKNOWLEDGMENTS
We would like to thank Drs. Raymond G. Goodwin and Jacques J. Peschon,
Immunex Corporation, Seattle, WA, for the TNFR KO mice used in these
experiments. In addition, the assistance of Ms. Victoria Lappi and Dr. Arindam
Bhattacharjee with RPA was greatly appreciated. This work was supported
through NIH grants ES04348 (LBS) and ES08395 (MSR.). TLH was supported in part by a USDA National Needs Graduate Fellowship.
REFERENCES
Akerman, P., Cote, P., Yang, S. Q., McClain, C., Nelson, S., Babgy, G. J., and
Diehl, A. M. (1992). Antibodies to tumor necrosis factor-␣ inhibit liver
regeneration after partial hepatectomy. Am. J. Physiol. 263, G579 –G585.
Bhattacharjee, A., Lappi, V. R., Rutherford, M. S., and Schook, L. B. (1998).
Molecular dissection of dimethylnitrosamine (DMN)-induced hepatotoxicity by mRNA differential display. Toxicol. Appl. Pharmacol. 150, 186 –195.
Blazka, M. E., Elwell, M. R., Holladay, S. D., Wilson, R. E., and Luster, M. I.
(1996). Histopathology of acetaminophen-induced liver changes: Role of
interleukin-1␣ and tumor necrosis factor-␣. Toxicol. Pathol. 24, 181–189.
Blazka, M. E., Wilmer, J. L., Holladay, S. D., Wilson, R. E., and Luster, M. I.
(1995). Role of proinflammatory cytokines in acetaminophen hepatotoxicity. Toxicol. Appl. Pharmacol. 133, 43–52.
Boess, F., Bopst, M., Althaus, R., Polsky, S., Cohen, S. D., Eugster, H.-P., and
Boelsterli, U. A. (1998). Acetaminophen hepatotoxicity in tumor necrosis
factor/lymphotoxin-␣ gene knockout mice. Hepatology 27, 1021–1029.
Brakebusch, C., Nophar, Y., Kemper, O., Engelmann, H., and Wallach, D.
(1992). Cytoplasmic truncation of the p55 tumor necrosis factor (TNF)
receptor abolishes signaling, but not induced shedding of the receptor.
EMBO J. 11, 943–950.
Bruccoleri, A., Gallucci, R., Germolec, D. R., Blackshear, P., Simeonova, P.,
Thurman, R. G., and Luster, M. I. (1997). Induction of early-immediate
genes by tumor necrosis factor ␣ contribute to liver repair following chemical-induced hepatotoxicity. Hepatology 25, 133–141.
Butler, W. H., and Hard, G. C. (1971). Hepatotoxicity of dimethylnitrosamine
in the rat with special reference to veno-occlusive disease. Exp. Mol. Pathol.
15, 209 –219.
Craddock, V. M. (1983). Nitrosamines and human cancer: Proof of an association? Nature 306, 638.
Czaja, M. J., Xu, J., and Alt, E. (1995). Prevention of carbon tetrachlorideinduced rat liver injury by soluble tumor necrosis factor receptor. Gastroenterology 108, 1849 –1854.
D’Sa-Eipper, C., Subramanian, T., and Chinnadurai, G. (1996). bfl-1, a bcl-2
homologue, suppresses p53-induced apoptosis and exhibits potent cooperative transforming activity. Cancer Res. 56, 3879 –3882.
Edwards, C. K., III, Borcherding, S. M., Zhang, J., and Borcherding, D. R.
(1994). Role of tumor necrosis factor ␣ in acute and chronic inflammatory
responses: novel therapeutic approaches. In Xenobiotics and Inflammation
(L. B. Schook, and D. L. Laskin, Eds.), pp. 97–147. Academic Press, Boca
Raton.
Engelmann, H., Holtmann, H., Brakebusch, C., Avni, Y. S., Sarov, I., Nophar,
Y., Hadas, E., Leitner, O., and Wallach, D. (1990). Antibodies to a soluble
form of a tumor necrosis factor (TNF) receptor have TNF-like activity.
J. Biol. Chem. 265, 14497–14504.
Espevik, T., Brockhaus, M., Loetscher, H., Nonstad, U., and Shalaby, R.
(1990). Characterization of binding and biological effects of monoclonal
antibodies against a human tumor necrosis factor receptor. J. Exp. Med. 171,
415– 426.
Ghezzi, P., Saccardo, B., and Bianchi, M. (1986). Recombinant tumor necrosis
factor depresses cytochrome P450-dependent microsomal drug metabolism
in mice. Biochem. Biophys. Res. Commun. 136, 316 –321.
Grell, M., Becke, F. M., Wajant, H., Mannel, D. N., and Scheurich, P. (1998).
TNF receptor type 2 mediates thymocyte proliferation independently of
TNF receptor type 1. Eur. J. Immunol. 28, 257–263.
Guengerich, F. P. (1987). Mammalian Cytochromes P-450. CRC Press, Boca
Raton.
Guengerich, F. P., Kim, D. H., and Iwasaki, M. (1991). Role of human
cytochrome P-450 IIE1 in the oxidation of many low molecular weight
cancer suspects. Chem. Res. Toxicol. 4, 168 –179.
Haggerty, H. G., and Holsapple, M. P. (1990). Role of metabolism in dimethylnitrosamine-induced immunosuppression: A review. Toxicology 63, 1–23.
Hohmann, H.-P., Remy, R., Pöschl, B., and van Loon, A. P. G. M. (1990).
Tumor necrosis factors-␣ and -␤ bind to the same two types of tumor
necrosis factor receptors and maximally activate the transcription factor
NF-␬B at low receptor occupancy and within minutes after receptor binding.
J. Biol. Chem. 265, 15183–15188.
Horn, T. L., Rutherford, M. S, and Schook, L. B. (1998). TNF-␣ is not required
for hepatic metallothionein expression during endotoxin or dimethylnitrosamine exposure in vivo. Toxicologist 42, 365.
Karsan, A., Yee, E., and Harlan, J. M. (1996a). Endothelial cell death induced
by tumor necrosis factor-␣ is inhibited by the Bcl-2 family member, A1.
J. Biol. Chem. 271, 27201–27204.
Karsan, A., Yee, E., Kaushansky, K., and Harlan, J. M. (1996b). Cloning of a
human Bcl-2 homologue: Inflammatory cytokines induce human A1 in
cultured endothelial cells. Blood 87, 3089 –3096.
Kayama, F., Yoshida, T., Elwell, M. R., and Luster, M. I. (1995). Role of
tumor necrosis factor-␣ in cadmium-induced hepatotoxicity. Toxicol. Appl.
Pharmacol. 131, 224 –234.
Kimura, T., Sakaida, I., Terai, S., Matsumura, Y., Uchida, K., and Okita, K.
(1997). Inhibition of tumor necrosis factor-␣ production retards liver regeneration after partial hepatectomy in rats. Biochem. Biophys. Res. Commun.
231, 557–560.
Koop, D. R. (1992). Oxidative and reductive metabolism by cytochrome P450
2E1. FASEB J. 6, 724 –730.
Kruppa, G., Thoma, B., Machleidt, T., Wiegmann, K., and Krönke, M. (1992).
Inhibition of tumor necrosis factor (TNF)-mediated NF-⌲B activation by
selective blockade of the human 55-kDa TNF receptor. J. Immunol. 148,
3152–3157.
TNF␣-MEDIATED HEPATOTOXIC RESPONSES
Lee, S. S. T., Buters, J. T. M., Pineau, T., Fernandez-Salguero, P., and
Gonzalez, F. J. (1996). Role of CYP2E1 in the hepatotoxicity of acetaminophen. J. Biol. Chem. 271, 12063–12067.
Leist, M., Gantner, F., Jilg, S., and Wendel, A. (1995). Activation of the
55-kDa TNF receptor is necessary and sufficient for TNF-induced liver
failure, hepatocyte apoptosis, and nitrite release. J. Immunol. 154, 1307–
1316.
Leist, M., Gantner, F., Naumann, H., Bluethmann, H., Vogt, K., BrigeliusFlohé, R., Nicotera, P., Volk, H.-D., and Wendel, A. (1997). Tumor necrosis
factor-induced apoptosis during the poisoning of mice with hepatotoxins.
Gastroenterology 112, 923–934.
Lockwood, J. F., Myers, M. J., Rutherford, M. S., and Schook, L. B. (1991).
Transcriptional changes in macrophage TNF-␣ expression following dimethylnitrosamine exposure in vivo. Immunopharmacology 22, 27–38.
Lockwood, J. F., Rutherford, M. S., Myers, M. J., and Schook, L. B. (1994).
Induction of hepatic acute-phase protein transcripts: Differential effects of
acute and subchronic dimethylnitrosamine exposure in vivo. Toxicol. Appl.
Pharmacol. 125, 288 –295.
Louis, H., Van Laethem, J.-L., Wu, W., Quertinmont, E., Degraef, C., Van den
Berg, K., Demols, A., Goldman, M., Le Moine, O., Geerts, A., and Devière,
J. (1998). Interleukin-10 controls neutrophilic infiltration, hepatocyte proliferation, and liver fibrosis induced by carbon tetrachloride in mice. Hepatology 28, 1607–1615.
Montesano, R., and Bartsch, H. (1976). Mutagenic and carcinogenic N-nitroso
compounds: Possible environmental hazards. Mutat. Res. 32, 179 –228.
Moos, A. B., Baecher-Steppan, L., and Kerkvliet, N. I. (1994). Acute inflammatory response to sheep red blood cells in mice treated with 2,3,7,8tetrachlorodibenzo-p-dioxin: The role of proinflammatory cytokines, IL-1
and TNF. Toxicol. Appl. Pharmacol. 127, 331–335.
Myers, M. J., and Schook, L. B. (1996). Immunotoxicity of nitrosamines. In
Experimental Immunotoxicology (R. J. Smialowicz, and M. P. Holsapple,
Eds.), pp. 351–366. CRC Press, Boca Raton.
Naume, B., Shalaby, R., Lesslauer, W., and Espevik, T. (1991). Involvement
of the 55- and 75-kDa tumor necrosis factor receptors in the generation of
lymphokine-activated killer cell activity and proliferation of natural killer
cells. J. Immunol. 146, 3045–3048.
Neumann, B., Machleidt, T., Lifka, A., Pfeffer, K., Vestweber, D., Mak, T. W.,
Holzmann, B., and Krönke, M. (1996). Crucial role of 55-kilodalton TNF
receptor in TNF-induced adhesion molecule expression and leukocyte organ
infiltration. J. Immunol. 156, 1587–1593.
Peschon, J. J., Torrance, D. S., Stocking, K. L., Glaccum, M. B., Otten, C.,
Willis, C. R., Charrier, K., Morrissey, P. J., Ware, C. B., and Mohler, K. M.
(1998). TNF receptor-deficient mice reveal divergent roles for p55 and p75
in several models of inflammation. J. Immunol. 160, 943–952.
Ray, S. D., Sorge, C. L., Kamendulis, L. M., and Corcoran, G. B. (1992).
Ca ⫹⫹-activated DNA fragmentation and dimethylnitrosamine-induced hepatic necrosis: effects of Ca ⫹⫹-endonuclease and poly(ADP-ribose) polymerase inhibitors in mice. J. Pharmacol. Exp. Ther. 263, 387–394.
Rechnagel, R. O., and Glende, E. A., Jr. (1973). Carbon tetrachloride hepatotoxicity: An example of lethal cleavage. CRC Crit. Rev. Toxicol. 2, 263–297.
Renton, K. W., and Armstrong, S. G. (1994). Immune-mediated downregula-
273
tion of cytochrome-P450 and related drug metabolism. In Immunotoxicology
and Immunopharmacology, 2nd ed. (J. H. Dean, M. I. Luster, A. E. Munson,
and I. Kimber, Eds.), pp. 501–512. Raven Press, New York.
Reynolds, E. S., and Ree, H. J. (1971). Liver parenchymal cell injury: VII.
Membrane denaturation following carbon tetrachloride. Lab. Invest. 25,
269 –278.
Rolfe, M., James, N. H., and Roberts, R. A. (1997). Tumor necrosis factor-␣
(TNF␣) suppresses apoptosis and induces DNA synthesis in rodent hepatocytes: A mediator of the hepatocarcinogenicity of peroxisome proliferators? Carcinogenesis 18, 2277–2280.
Schook, L. B., Lockwood, J. F., Yang, S.-D., and Myers, M. J. (1992).
Dimethylnitrosamine (DMN)-induced IL-1␤, TNF-␣, and IL-6 inflammatory cytokine expression. Toxicol. Appl. Pharmacol. 116, 110 –116.
Shalaby, M. R., Sundan, A., Loetscher, H., Brockhaus, M., Lesslauer, W., and
Espevik, T. (1990). Binding and regulation of cellular functions by monoclonal antibodies against human tumor necrosis factor receptors. J. Exp.
Med. 172, 1517–1520.
Shi, J., Aisaki, K., Ikawa, Y., and Wake, K. (1998). Evidence of hepatocyte
apoptosis in rat liver after the administration of carbon tetrachloride. Am. J.
Pathol. 153, 515–525.
Takehara, T., Hayashi, N., Mita, E., Kanto, T., Tatsumi, T., Sasaki, Y.,
Kasahara, A., and Hori, M. (1998). Delayed Fas-mediated hepatocyte
apoptosis during liver regeneration in mice: Hepatoprotective role of TNF␣.
Hepatology 27, 1643–1651.
Tartaglia, L. A., and Goeddel, D. V. (1992). Two TNF receptors. Immunol.
Today 13, 151–153.
Tartaglia, L. A., Weber, R. F., Figari, I. S., Reynolds, C., Palladino, M. A., Jr.,
and Goeddel, D. V. (1991). The two different receptors for tumor necrosis
factor mediate distinct cellular responses. Proc. Natl. Acad. Sci. U S A 88,
9292–9296.
Thoma, B., Grell, M., Pfizenmaier, K., and Scheurich, P. (1990). Identification
of a 60-kD tumor necrosis factor (TNF) receptor as the major signaltransducing component in TNF responses. J. Exp. Med. 172, 1019 –1023.
Thomsen, M. S., Loft, S., Roberts, D. W., and Poulsen, H. E. (1995). Cytochrome P4502E1 inhibition by propylene glycol prevents acetaminophen
(paracetamol) hepatotoxicity in mice without cytochrome P4501A2 inhibition. Pharmacol. Toxicol. 76, 395–399.
Wang, B., Fujisawa, H., Zhuang, L., Kondo, S., Shivji, G. M., Kim, C. S., Mak,
T. W., and Sauder, D. N. (1997). Depressed Langerhans cell migration and
reduced contact hypersensitivity response in mice lacking TNF receptor p75.
J. Immunol. 159, 6148 – 6155.
Wang, B., Kondo, S., Shivji, G. M., Fujisawa, H., Mak, T. W., and Sauder,
D. N. (1996). Tumor necrosis factor receptor II (p75) signaling is required
for the migration of Langerhans’ cells. Immunology 88, 284 –288.
Warren, G. W., Poloyac, S. M., Gary, D. S., Mattson, M. P., and Blouin, R. A.
(1999). Hepatic cytochrome P-450 expression in tumor necrosis factor-␣
receptor (p55/p75) knockout mice after endotoxin administration. J. Pharmacol. Exp. Ther. 288, 945–950.
Yamada, Y., and Fausto, N. (1998). Deficient liver regeneration after carbon
tetrachloride injury in mice lacking type 1 but not type 2 tumor necrosis
factor receptor. Am. J. Pathol. 152, 1577–1589.