This information is current as
of June 15, 2017.
Activation of Caspase 3 (CPP32)-Like Proteases
Is Essential for TNF- α-Induced Hepatic
Parenchymal Cell Apoptosis and
Neutrophil-Mediated Necrosis in a Murine
Endotoxin Shock Model
Hartmut Jaeschke, Michael A. Fisher, Judy A. Lawson, Carol A.
Simmons, Anwar Farhood and David A. Jones
J Immunol 1998; 160:3480-3486; ;
http://www.jimmunol.org/content/160/7/3480
Subscription
Permissions
Email Alerts
This article cites 43 articles, 18 of which you can access for free at:
http://www.jimmunol.org/content/160/7/3480.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 1998 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
References
Activation of Caspase 3 (CPP32)-Like Proteases Is Essential
for TNF-a-Induced Hepatic Parenchymal Cell Apoptosis and
Neutrophil-Mediated Necrosis in a Murine Endotoxin Shock
Model1
Hartmut Jaeschke,2* Michael A. Fisher,* Judy A. Lawson,* Carol A. Simmons,*
Anwar Farhood,† and David A. Jones3*
L
iver dysfunction and failure are common problems during
endotoxemia and sepsis (1). It is generally accepted that
proinflammatory cytokines, especially TNF-a, are critical
for the pathophysiology (2–5). TNF-a has a wide variety of effects
on leukocytes and liver cells that support inflammation; e.g.,
TNF-a is one of the mediators that induce up-regulation of the b2
integrin Mac-1 (CD11b/CD18) on neutrophils during endotoxemia
in vivo (6); TNF-a also primes Kupffer cells (7) and neutrophils
(8) for release of cytotoxic mediators. In addition, TNF-a activates
the transcription factor NF-kB4 in endothelial cells and hepato-
*Department of Pharmacology, Pharmacia & Upjohn, Inc., Kalamazoo, MI 49007;
and †Department of Pathology, University of Texas Health Science Center, Houston,
TX 77030
Received for publication September 9, 1997. Accepted for publication December
1, 1997.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
The work was supported in part by National Institutes of Health Grant RO1 ES
06091.
2
Address correspondence and reprint requests to Dr. Hartmut Jaeschke, Department
of Pharmacology (7250-300-210), Pharmacia & Upjohn, Inc., 301 Henrietta Street,
Kalamazoo, MI 49007. E-mail address: [email protected]
3
Current address: Huntsman Cancer Institute, Salt Lake City, UT 84112.
Abbreviations used in this paper: NF-kB, nuclear factor kB; ET, endotoxin; Gal,
galactosamine; ALT, alanine aminotransferase; TUNEL, terminal deoxynucleotidyl
transferase-mediated dUTP nick end labeling; ICE, IL-1b-converting enzyme;
Z-VAD, Z-Val-Ala-Asp-CH2F; Ac-DEVD-MCA, acetyl-Asp-Glu-Val-Asp-(4-methylcoumaryl-7-amide); Ac-YVAD-MCA, acetyl-Tyr-Val-Ala-Asp-(4-methylcoumaryl-7-amide); Ac-DEVD-Ald, acetyl-Asp-Glu-Val-Asp-CHO); Ac-YVAD-Ald,
4
Copyright © 1998 by The American Association of Immunologists
cytes during endotoxemia (9), which leads to the transcriptional
activation of a number of proinflammatory genes, e.g., chemokines
(10, 11), nitric oxide synthase (12, 13), and adhesion molecules
such as ICAM-1 (14, 15), VCAM-1 (16), and selectins (9, 15).
Therefore, TNF-a alone or in combination with IL-1 and complement is responsible for neutrophil sequestration in hepatic sinusoids during endotoxemia and sepsis (14, 17). After transmigration, neutrophils attack parenchymal cells and cause severe liver
cell necrosis (14, 16, 18). Thus, TNF-a is a critical early mediator
for an acute inflammatory response in the liver during endotoxemia and sepsis.
TNF-a can act through binding to two different cell surface
receptors, i.e., the 55-kDa TNF-R1 and the 75-kDa TNF-R2 (19,
20). Most of the proinflammatory effects of TNF-a are mediated
through TNF-R1 (20). However, TNF-a can also induce apoptosis
through the death domain of TNF-R1 (21). Recently, hepatocellular apoptosis has been characterized in various models of endotoxemia (22–25), and it was confirmed that this effect was mediated in vivo through TNF-R1 (25, 26). In all eukaryotic cells, the
intracellular signal transduction pathway leading to apoptosis involves the activation of a cascade of cysteine proteases (caspases/
ICE proteases) (27–29). Currently, there are 10 human caspases
identified (30) with equivalent enzymes in the mouse (31). The
activation of caspase 3 (CPP32)-like proteases in liver cells was
observed during the development of apoptosis after various insults,
i.e., anti-Fas Ab in vivo (32), TGF-b1 (33), staurosporine (33), and
acetyl-Tyr-Val-Ala-Asp-CHO; CrmA, cowpox viral serpin cytokine response modifier A; Ac-YVAD-cmk, acetyl-Tyr-Val-Ala-Asp-chloromethylketone.
0022-1767/98/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Endotoxin (ET)-induced liver failure is characterized by parenchymal cell apoptosis and inflammation leading to liver cell necrosis. Members of the caspase family have been implicated in the signal transduction pathway of apoptosis. The aim of this study
was to characterize ET-induced hepatic caspase activation and apoptosis and to investigate their effect on neutrophil-mediated
liver injury. Treatment of C3Heb/FeJ mice with 700 mg/kg galactosamine (Gal) and 100 mg/kg Salmonella abortus equi ET
increased caspase 3-like protease activity (Asp-Val-Glu-Asp-substrate) by 1730 6 140% at 6 h. There was a parallel enhancement
of apoptosis (assessed by DNA fragmentation ELISA and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling
assay). In contrast, activity of caspase 1 (IL-1b-converting enzyme)-like proteases (Tyr-Val-Ala-Asp-substrate) did not change
throughout the experiment. Caspase 3-like protease activity and apoptosis was not induced by Gal/ET in ET-resistant mice
(C3H/HeJ). Furthermore, only murine TNF-a but not IL-1ab increased caspase activity and apoptosis. Gal/ET caused neutrophildependent hepatocellular necrosis at 7 h (area of necrosis, 45 6 3%). Delayed treatment with the caspase 3-like protease inhibitor
Z-Val-Ala-Asp-CH2F (Z-VAD) (10 mg/kg at 3 h) attenuated apoptosis by 81 to 88% and prevented liver cell necrosis (<5%).
Z-VAD had no effect on the initial inflammatory response, including the sequestration of neutrophils in sinusoids. However,
Z-VAD prevented neutrophil transmigration and necrosis. Our data indicate that activation of the caspase 3 subfamily of cysteine
proteases is critical for the development of parenchymal cell apoptosis. In addition, excessive hepatocellular apoptosis can be an
important signal for transmigration of primed neutrophils sequestered in sinusoids. The Journal of Immunology, 1998, 160:
3480 –3486.
The Journal of Immunology
Materials and Methods
Animals
Male mice, strains C3Heb/FeJ (ET-sensitive) and C3H/HeJ (ET-resistant)
(20 –25 g body weight), were purchased from The Jackson Laboratory (Bar
Harbor, ME). The animals had free access to food (certified rodent diet no.
5002C, PMI Feeds, Richmond, IN) and water. The experimental protocols
followed the criteria of Pharmacia & Upjohn (Kalamazoo, MI) and of the
National Research Council for the care and use of laboratory animals in
research. Animals were treated i.p. with 700 mg/kg D-Gal (Sigma Chemical
Co., St. Louis, MO) and 100 mg/kg Salmonella abortus equi ET (Sigma
Chemical) dissolved in sterile PBS (pH 7.0). Some animals were treated
with 3 3 10 mg/kg of the caspase inhibitor Z-VAD (Enzyme Systems
Products, Dublin, CA); the drug was injected 3, 4.5, and 5.5 h after Gal/ET
administration. Vehicle control animals received DMSO (1 ml/kg) at the
same time. Other experiments included i.v. injection of murine recombinant TNF-a (15 mg/kg; specific activity, 4 3 108 U/mg) (Genzyme, Cambridge, MA), murine rIL-1a (13 mg/kg; specific activity, 8 3 106 U/mg)
(Genzyme), or murine rIL-1b (23 mg/kg; specific activity, 1.5 3 106 U/mg)
(Genzyme) in Gal-sensitized animals.
Experimental protocols
The animals (5– 8 per treatment group) were killed by cervical dislocation
various times after administration of Gal/ET or a cytokine (TNF-a, IL-1a,
IL-1b). Blood was collected from the right ventricle into a heparinized
syringe and centrifuged, and plasma was used for determination of ALT
activity with Sigma test kit DG 159-UV. Pieces of the liver were immediately homogenized for caspase activity measurements; other parts of each
liver were frozen in liquid nitrogen and stored at 280°C for analysis of
DNA fragmentation, fixed in phosphate-buffered formalin for histologic
analysis, or embedded in OCT embedding medium (Miles Diagnostic Division, Elkhart, IN), and snap-frozen in methylbutane cooled in liquid nitrogen for immunohistochemistry.
Caspase activity
Freshly excised liver was homogenized in 25 mM HEPES buffer (pH 7.5)
containing 5 mM EDTA, 2 mM DTT, and 0.1% CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate). After centrifugation at
14,000 3 g, the diluted supernatant was assayed for caspase activity using
synthetic fluorogenic substrates: Ac-DEVD-MCA (Ac-Asp-Glu-Val-AspMCA) (Peptide Institute, Inc., Osaka, Japan) for caspase 3 (CPP32)/
caspase 7 (Mch3) and Ac-YVAD-MCA (Ac-Tyr-Val-Ala-Asp-MCA)
(Peptide Institute) for caspase 1 (ICE) at concentrations of 50 mM. The
kinetics of the proteolytic cleavage of the substrates was monitored in a
fluorescence microplate reader (Fmax; Molecular Devices, Corp., Sunnyvale, CA) using an excitation wavelength of 360 nm and an emission wavelength of 460 nm. The fluorescence intensity was calibrated with standard
concentrations of MCA, and the caspase activity was calculated from the
slope of the recorder trace and expressed in picomols per minute per mg of
protein. Protein concentrations in the supernatant were assayed using the
bicinchoninic acid kit (Sigma). For the inhibitor studies in vitro, AcDEVD-Ald, Ac-YVAD-Ald (Peptide Institute), or CrmA (Kamiya Biomedical Co., Seattle, WA) were added to the supernatant (3.3 nM-10 mM)
15 min before adding the substrate Ac-DEVD-MCA.
Histology
Formalin-fixed portions of the liver were embedded in paraffin and 5-mmthick sections were cut. Neutrophils were stained by the AS-D chloroacetate esterase technique as described in detail (17). Neutrophils were identified by positive staining and morphology and were counted in 50 high
power fields (3400) using a Nikon Labophot microscope (Nikon, Melville,
NY). Only those neutrophils present within sinusoids or extravasated into
the tissue were counted; generally, there are no neutrophils present in large
hepatic vessels, e.g., venules, at that time (18). The percentage of necrotic
area was estimated by evaluating parallel sections stained with hematoxylin and eosin. The pathologist (A.F.) performing the histologic evaluation
(polymorphonuclear leukocytes, area of necrosis) was blinded as to the
treatment of animals.
Apoptosis assays
Apoptotic cells were determined using an ApopTag in situ apoptosis peroxidase detection kit (terminal deoxynucleotidyl transferase-mediated
dUTP nick end labeling; TUNEL assay) (Oncor, Gaithersburg, MD) and a
MicroProbe staining station (Fisher Scientific, Pittsburgh, PA). Frozen
liver samples were cryosectioned (6 mm; two sections/sample, 150 mm
apart) using a Leitz 1720 cryostat (W. Nuhsbaum, McHenry, IL) and assayed for direct immunoperoxidase detection of digoxigenin-labeled
genomic DNA. Staining was conducted according to the manufacturer’s kit
protocol for fresh frozen tissue samples. DNase 1 treatment of additional
sections, to nick all DNA, served as positive controls at a concentration of
1 mg/ml determined after a preliminary dilution experiment (Appligene
DNase 1, Oncor). This was prepared and conducted according to manufacturer’s instructions. TUNEL-stained mouse liver tissue sections were
quantitated for apoptotic cells using computer-assisted image analysis and
Optimas software (Bioscan, Edmonds, WA). Visual images from TUNELstained slides were captured with a digital camera attached to a microscope
with modifications of previously published methods (38). A color threshold
was set (dark brown for diaminobenzidine) to match visual recognition of
apoptotic cells. Ten random microscopic regions of interest (each covering
a surface area of 0.55 mm2) were evaluated per tissue slide (two slides/
sample) and for the number of apoptotic cells captured into computer memory and downloaded into Excel 4.0 (Microsoft Co, Redmond, WA) spreadsheets for statistical analysis.
For the cell death detection ELISA (Boehringer Mannheim, Indianapolis, IN), a 20% homogenate in 50 mM sodium phosphate buffer (120 mM
NaCl, 10 mM EDTA) was prepared and centrifuged at 14,000 3 g. Diluted
supernatant was used for the ELISA. In this test, the kinetics of product
generation (Vmax) is a measure of DNA fragmentation. The Vmax values
obtained for untreated controls (100%) are compared with those in treated
groups. The assay allows the specific quantitation of histone-associated
DNA fragments (mono- and oligonucleosomes) in the cytoplasmic fraction
of cell lysates and was used extensively to demonstrate apoptosis in the
liver in vivo (22, 23, 26).
Isolation of mouse liver cells
Animals (n 5 4 per group) were anesthetized with a ketamine mixture (225
mg/kg ketamine; 11.4 mg/kg xylazine; 2.3 mg/kg acepromazine) i.m. The
liver was perfused free of blood in an open system for 5 to 10 min using
an oxygenated Ca21-free Hanks’ buffer. A collagenase-supplemented (25
mg/100 ml buffer) Hanks’ buffer was used to digest the liver. When good
digestion was obtained (;10 min), the liver was removed, minced, and
strained through a tissue sieve. Cells were then centrifuged at 50 3 g for
3 min. The supernatant (nonparenchymal cells) was removed and saved.
The pellet (parenchymal cells) was resuspended in Hanks’ buffer and spun
at 50 3 g for 3 min. The supernatant was combined with the supernatant
from the first spin, and the pellet was resuspended. Cell fractions were then
spun at 600 3 g for 10 min. The supernatants were discarded, and the
nonparenchymal pellet was resuspended in pronase buffer (200 mg/50 ml
buffer) and stirred for 10 min to remove any hepatocytes in the suspension.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
hypoxia-reoxygenation (34) in vitro. Moreover, inhibitors of
caspases are highly effective in preventing apoptotic cell death
(32–34). This suggests that caspase inhibitors may allow blockage
of apoptotic cell injury without affecting the TNF-a-induced proinflammatory signal transduction pathway.
Although apoptotic and necrotic hepatocytes were identified
during endotoxemia (14, 15, 18, 22–25, 35), the pathophysiologic
significance of hepatocellular apoptosis for the later development
of liver cell necrosis remains unclear. Previous interventions directed against TNF-a that protected against ET-induced liver injury, e.g., neutralizing Abs, TNF receptor knockout mice, and inhibition of TNF-a gene transcription and synthesis (2, 14, 25, 26,
36, 37), inhibited both the inflammatory and the apoptotic response
and did not allow a conclusion regarding the relationship between
apoptotic cell death and necrosis in the liver. The beneficial effects
observed with selective blockade of adhesion molecules on neutrophils (35) and liver cells (14, 16) suggested a critical role for
neutrophils in the development of severe necrosis. Thus, apoptosis
under these conditions may be an epiphenomenon of limited relevance or could be an important signal for neutrophil transmigration. To address this critical question, we used the Gal/ET shock
model with its extensively described inflammatory (14, 16, 18, 35,
37) and apoptotic (22, 24, 26) response to characterize the role of
caspase activation in hepatocellular apoptosis and, by using
caspase inhibitors to selectively prevent apoptosis, to study the
relevance of apoptosis for the neutrophil-induced liver cell
necrosis.
3481
3482
This solution was then spun at 600 3 g for 10 min, and the pellet was
washed once. Both cell fractions were exposed to an ammonium chloride
lysing solution for 10 min to lyse contaminating RBC. Cells were washed
again, resuspended, and counted. Cell fractions were .98% pure and
.95% viable as judged by trypan blue exclusion. Cell concentrations were
adjusted to 4 3 106 cells/ml with either caspase assay buffer or 50 mM
phosphate buffer (DNA fragmentation ELISA).
Statistics
All data are given as mean 6 SE. Statistical significance between the
control group and a treated group was determined with the unpaired Student’s t test, or Wilcoxon rank sum test. Comparisons between multiple
groups were performed with one way ANOVA followed by the Bonferroni
t test. p , 0.05 was considered significant.
Results
Administration of Gal/ET caused a time-dependent increase of hepatic caspase activity using Ac-DEVD-MCA as a substrate (Fig.
1A). In contrast, no increased protease activity was detected with
the caspase 1 substrate Ac-YVAD-MCA. These results suggest the
activation of proapoptotic caspase 3-like proteases during endotoxemia in Gal-sensitized animals. Parallel to the increase of
caspase 3-like protease activity, there was enhanced DNA fragmentation, indicating cells undergoing apoptosis at the same time
(Fig. 1B). Neither Gal nor ET alone caused an increase in hepatic
caspase activity or DNA fragmentation (data not shown). Apoptosis in the liver of Gal/ET-treated animals was confirmed using the
TUNEL assay (Fig. 2). Quantitative analysis showed few apoptotic
cells in controls (about 0.2% of hepatocytes) and no significant
change up to 4 h after Gal/ET treatment (Fig. 3). However, at 5 and
6 h, the number of apoptotic cells increased dramatically reaching
13% at 6 h (65-fold increase). These data indicate that the increase
of proapoptotic caspase activity correlated with the increase of
apoptotic cell death in this model. Because the intact tissue did not
allow us to distinguish which liver cell types had been affected,
parenchymal cells (hepatocytes) and nonparenchymal cells (mixture of Kupffer cells and endothelial cells) were isolated 5.5 h after
Gal-ET administration. Measurement of DEVD-MCA caspase activity and DNA fragmentation showed significant increases only in
hepatocytes but not in the nonparenchymal cell fraction (Fig. 4).
To further characterize the type of caspase involved, inhibitor
studies were performed using liver homogenate of Gal/ET-treated
animals (6 h). The inhibitor of caspase 3-like proteases, AcDEVD-Ald, dose dependently inhibited the enzyme activity in the
liver (Fig. 5). An IC50 of 16.75 nM was calculated from these data.
In contrast, the caspase 1 inhibitor Ac-YVAD-Ald or CrmA, an
effective inhibitor of caspase 1 and the proapoptotic caspase 8 had
no or very limited effects on Gal/ET-induced protease activity in
the liver. The extrapolated IC50 for YVAD-Ald was 19.6 mM.
To identify the critical mediators involved in the activation of
caspases in vivo, ET-resistant animals, which do not generate cytokines in response to ET, were treated with Gal/ET. In contrast to
the sensitive animals, ET-resistant animals did not show enhanced
caspase 3-like protease activity or evidence for increased DNA
fragmentation (Table I). Injection of TNF-a, IL-1a, or IL-1b in
combination with Gal indicated that only Gal/TNF-a was able to
induce an increase in caspase 3-like protease activity and apoptosis
in ET-resistant (Table I) and in ET-sensitive animals (data not
shown). These findings suggest that ET-induced proapoptotic
caspase activation and apoptotic cell death in hepatocytes is mediated by TNF-a.
Because we previously provided evidence that neutrophils are
critical for ET-induced hepatocellular necrosis in this model, we
tested the hypothesis that caspase activation and apoptosis are relevant for the inflammatory response. A large group of animals was
treated with Gal/ET. After 3 h, i.e., at a time when TNF-a has been
generated and TNF-a-mediated inflammatory responses were initiated (NF-kB activation, transcription of adhesion molecules) (2,
6, 9, 14, 16, 18), animals either were left untreated (disease control) or were injected with the caspase inhibitor Z-VAD in DMSO
(10 mg/kg) or with DMSO alone (1 ml/kg). Z-VAD or vehicle
treatment was repeated twice at 90-min intervals. Gal/ET administration caused severe liver injury at 7 h as indicated by a significant increase in plasma ALT activities and widespread hepatocellular necrosis (45% of hepatocytes) (Fig. 6A). Z-VAD, in contrast
to the vehicle DMSO, effectively inhibited liver injury. Necrotic
cell injury was almost eliminated; i.e., only few individual hepatocytes could be identified as necrotic. In contrast, Z-VAD did not
reduce the overall number of neutrophils in the liver (Fig. 6B).
However, virtually all neutrophils (.95%) remained localized
within sinusoids. In the vehicle group, 30 to 35% of neutrophils
were extravasated. To demonstrate the effect of Z-VAD treatment
on caspase activity and apoptosis, the experiment was repeated,
and the animals were killed at 6 h (before the onset of massive
hemorrhage and necrosis). Z-VAD completely inhibited caspase
3-like protease activity in the liver but had no effect on the baseline
caspase 1 activity (Fig. 7). Z-VAD treatment also significantly
reduced the increase in apoptotic cell death as determined by two
independent criteria, DNA fragmentation (81% inhibition) and
TUNEL assay (88% reduction of apoptotic cells) (Table II).
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 1. Time course of hepatic caspase activity (A) and DNA fragmentation (B) as an indicator for apoptosis after i.p. administration of 700
mg/kg Gal and 100 mg/kg S. abortus equi endotoxin (Gal/ET). Caspase
activity was determined by measuring the proteolytic cleavage of the
caspase 1 substrate Ac-YVAD-MCA and the caspase 3 substrate AcDEVD-MCA. 100% values are 30.6 6 0.9 pmol MCA/min/mg protein
(DEVD) and 118.2 6 5.0 pmol MCA/min/mg protein (YVAD). DNA fragmentation was measured by ELISA; values are reported as % changes of
Vmax values obtained from controls (5100%). Data represent means 6 SE
of six to eight animals per time point. n.d., not determined in tissue with
severe hemorrhage because of interferences with the assay. *, p , 0.05
(compared with t 5 0).
CASPASE ACTIVATION AND ET-INDUCED LIVER INJURY
The Journal of Immunology
3483
Discussion
The principal objective of this investigation was to study the role
of caspase activation in hepatocellular apoptosis and neutrophilinduced necrosis during ET-mediated liver injury in vivo. Our data
demonstrated a substantial increase of caspase activity with
DEVD-MCA, a substrate of caspase 3-like proteases.
However, no increase of activity was seen with YVAD-MCA, a
substrate more specific for the caspase 1 subfamily. Furthermore,
the enhanced caspase activity could be inhibited by DEVD-Ald but
FIGURE 3. Quantitative analysis of apoptotic cells. TUNEL-stained
mouse liver tissue sections were quantitated for apoptotic cells using computer-assisted image analysis as described in Materials and Methods. Cells
were counted in 20 random fields (3125) of 2 slides from each liver and
expressed as apoptotic cells per field. Data represent means 6 SE of six to
eight animals. *, p , 0.05 (compared with t 5 0).
not by YVAD-Ald or CrmA. These data suggest that in the Gal/ET
model in vivo, there is a predominant activation of the caspase
3-like subfamily of proteases. Caspase activation correlated with
development of hepatocellular apoptosis in vivo. These results are
consistent with critical involvement of caspase 3 subfamily members in apoptosis in a variety of cell types (27–29). The effect of
FIGURE 4. Caspase activity and DNA fragmentation in parenchymal
cells (hepatocytes) and nonparenchymal cells in untreated controls and 6 h
after i.p. administration of 700 mg/kg Gal and 100 mg/kg S. abortus equi
endotoxin (Gal/ET). Caspase activity was determined by measuring the
proteolytic cleavage of the caspase 3 substrate Ac-DEVD-MCA. 100%
value: 24.6 6 4.5 pmol of MCA/min/mg protein (DEVD). DNA fragmentation was measured by ELISA; values are reported as percent changes of
vmax obtained from controls (5100%). Data represent means 6 SE of four
animals per group. *, p , 0.05 (compared with control).
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 2. Detection of hepatocellular DNA fragmentation using the TUNEL method as described in Materials and Methods. Arrows indicate apoptotic
cells. A, Untreated control; B, liver section of an ET-sensitive mouse 6 h after treatment with Gal/ET; C, liver section of an ET-resistant mouse 6 h after
Gal/ET treatment; D, a liver section that was pretreated with DNase 1 to nick all DNA, served as positive control. All samples: 3312.5.
3484
CASPASE ACTIVATION AND ET-INDUCED LIVER INJURY
FIGURE 5. Dose-dependent inhibition of caspase activity (Ac-DEVDMCA) in liver homogenate derived from animals treated with 700 mg/kg
Gal/100 mg/kg ET for 6 h. Inhibitors (CrmA, Ac-DEVD-Ald, or AcYVAD-Ald) were added to the supernatant 15 min before the substrate.
FIGURE 6. Liver injury, as assessed by plasma ALT activities and histologically by the area of necrosis (A), and hepatic neutrophil sequestration
(B) were evaluated in C3Heb/FeJ mice before (controls, C) and 7 h after
the combined administration of 700 mg/kg Gal and 100 mg/kg S. abortus
equi endotoxin (G/E). Some animals received 3 3 10 mg/kg of the caspase
inhibitor Z-VAD; the drug was injected 3, 4.5, and 5.5 h after G/E administration. Vehicle control animals received DMSO (1 ml/kg) at the same
time. Data represent means 6 SE of eight animals per group. *, p , 0.05
(compared with control); #, p , 0.05 (ZVAD vs DMSO).
showed that ;13% of hepatocytes could be identified as undergoing apoptosis and only ,10% of the cells were necrotic at 6 h after
Gal/ET treatment. One h later, 38 to 45% of hepatocytes were
Table I. Caspase activation and apoptosis in ET-resistant animalsa
DEVD
YVAD
DNA Fragmentation
(pmol/min/mg protein) (pmol/min/mg protein)
Vmax (%)
Controls
Gal/ET
Gal/TNF-a
Gal/IL-1a
Gal/IL-1b
28.7 6 3.0
25.5 6 4.7
190.3 6 44.5*
34.1 6 4.5
34.6 6 6.0
86.2 6 4.3
100.6 6 5.5
92.1 6 2.7
99.6 6 6.6
101.0 6 4.1
100 6 5
78 6 13
410 6 5*
88 6 13
83 6 19
a
Hepatic caspase activity and DNA fragmentation (% of control) as indicator for
apoptosis were determined in C3H/HeJ (ET-resistant) mice 5 h after administration of
700 mg/kg Gal in combination with 100 mg/kg S. abortus equi ET (Gal/ET), 15 mg/kg
murine recombinant TNF-a (Gal/TNF-a), 13 mg/kg murine rIL-1a (Gal/IL-1a), or 23
mg/kg murine rIL-1b (Gal/IL-1b). DNA fragmentation was measured by ELISA;
values are reported as % changes of Vmax obtained from controls (5100%). Caspase
activity was determined by measuring the proteolytic cleavage of the caspase 1 substrate Ac-YVAD-MCA and the caspase 3 substrate Ac-DEVD-MCA. Data represent
means 6 SE of five animals per group. *p , 0.05 (compared with controls).
FIGURE 7. Increase in hepatic caspase activity before (controls, C) and
6 h after i.p. administration of 700 mg/kg Gal and 100 mg/kg S. abortus
equi endotoxin (G/E). Caspase activity was determined by measuring the
proteolytic cleavage of the caspase 1 substrate Ac-YVAD-MCA and the
caspase 3 substrate Ac-DEVD-MCA. 100% values are 36.9 6 3.1 pmol of
MCA/min/mg protein (DEVD) and 72.5 6 4.8 pmol of MCA/min/mg protein (YVAD). Some animals were treated with 3 3 10 mg/kg of the caspase
inhibitor Z-VAD; the drug was injected 3, 4.5, and 5.5 h after G/E administration. Vehicle control animals received DMSO (1 ml/kg) at the same
time. Data represent mean 6 SE of six animals per group. *, p , 0.05
(compared with control); #, p , 0.05 (ZVAD vs DMSO).
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Gal/ET treatment on hepatic caspase activity was not observed in
ET-resistant animals that do not generate cytokines in response to
ET. Furthermore, caspase activation could be induced with TNF-a
but not with IL-1 in ET-resistant and -sensitive mice. These findings strongly indicate that Gal/ET-induced activation of caspase
3-like proteases in vivo and hepatocellular apoptosis is mediated
by TNF-a. Other stimuli that induced caspase activation in hepatocytes, e.g., Fas Ab (32), TGF-b1 (33), and hypoxia-reoxygenation (34), showed increased protease activity with DEVD-MCA
and YVAD-MCA as substrates. Inhibitors of caspase 1 (YVADAld, YVAD-cmk) and caspase 3 (DEVD-Ald, Z-VAD) inhibited
hepatocyte apoptosis in vitro (34, 39) and after Fas Ab administration in vivo (32, 39). On the basis of these data, Rouquet et al.
(40) suggested that at least two distinct pathways of Fas signaling
exist in hepatocytes. Activation of caspase 1- and caspase 3-like
proteases are required, but these pathways involve different subclasses of serine proteases and can be selectively modulated by
protein tyrosine kinase inhibitors (40). Interestingly, actinomycin/
TNF-a-induced apoptosis in isolated hepatocytes and in vivo
could be inhibited by YVAD-cmk (39). In contrast, our data indicated that Gal/TNF-a-mediated apoptosis correlated only with increased caspase 3-like activity and could only be inhibited with
predominantly caspase 3 inhibitors, e.g., DEVD-Ald. This would
suggest that there might also be multiple pathways for TNF-a
signaling of apoptosis.
Hepatocellular apoptosis has been described in several models
of ET shock (22–26); however, the importance of apoptosis for the
overall injury and organ failure in these models remained unclear.
Quantitative analysis of apoptotic and necrotic parenchymal cells
The Journal of Immunology
3485
Table II. Effect of the caspase inhibitor Z-VAD on Gal/ET-induced
apoptosisa
Controls
Gal/ET
Gal/ET 1 DMSO
Gal/ET 1 Z-VAD
DNA Fragmentation
(Vmax (%))
TUNEL Assay
(Apoptotic Cells/Field)
100 6 13
489 6 7*
515 6 11*
177 6 16*†
2.6 6 0.3
165.5 6 35.2*
197.8 6 44.5*
24.0 6 1.2*†
a
Evaluation of apoptotic cell injury by DNA fragmentation and the TUNEL assay
before (Controls) and 6 h after i.p. administration of 700 mg/kg Gal and 100 mg/kg
S. abortus equi ET (Gal/ET). Some animals were treated with 3 3 10 mg/kg of the
caspase inhibitor Z-VAD (Z-Val-Ala-Asp-CH2F); the drug was injected 3, 4.5, and
5.5 h after Gal/ET administration. Vehicle control animals received DMSO (1 ml/kg)
at the same time. TUNEL-stained mouse liver tissue sections were quantitated for
apoptotic cells using computer-assisted image analysis as described in Materials and
Methods. Cells were counted in 20 random fields (3125) of 2 slides from each liver
and expressed as apoptotic cells per field. DNA fragmentation was measured by
ELISA; values are reported as % changes of Vmax obtained from controls (5100%).
Data represent means 6 SE of six animals. *p , 0.05 (compared with controls);
†p , 0.05 (compared with Gal/ET or Gal/ET 1 DMSO).
References
1. Parrillo, J. E. 1993. Pathogenetic mechanisms of septic shock. N. Engl. J. Med.
328:1471.
2. Schlayer, H. J., H. Laaff, T. Peters, M. Woort-Menker, H. C. Estler, U. Karck,
H. E. Schaefer, and K. Decker. 1988. Involvement of tumor necrosis factor in
endotoxin-triggered neutrophil adherence to sinusoidal endothelial cells of mouse
liver and its modulation in acute phase. J. Hepatol. 7:239.
3. Wang, P., and I. H. Chaudry. 1996. Mechanism of hepatocellular dysfunction
during hyperdynamic sepsis. Am. J. Physiol. 270:R927.
4. Wang, P., Z. F. Ba, and I. H. Chaudry. 1997. Mechanism of hepatocellular dysfunction during early sepsis-key role of increased gene expression and release of
proinflammatory cytokines tumor necrosis factor and interleukin-6. Arch. Surg.
132:364.
5. Casey, L. C. 1993. Role of cytokines in the pathogenesis of cardiopulmonaryinduced multisystem organ failure. Ann. Thorac. Surg. 56:S92.
6. Witthaut, R., A. Farhood, C. W. Smith, and H. Jaeschke. 1994. Complement and
tumor necrosis factor-a contribute to Mac-1 (CD11b/CD18) up-regulation and
systemic neutrophil activation during endotoxemia in vivo. J. Leukocyte Biol.
55:105.
7. Bautista, A. P., A. Schuler, Z. Spolarics, and J. J. Spitzer. 1991. Tumor necrosis
factor-a stimulates superoxide anion generation by perfused rat liver and Kupffer
cells. Am. J. Physiol. 261:G891.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
necrotic. These data indicate that quantitively, apoptotic cell death
could not explain the three- to fourfold higher number of necrotic
cells 1 h later. Thus, apoptosis appears to be a trigger mechanism
for the more severe attack of neutrophils. This hypothesis was
investigated by the use of caspase inhibitors, which allowed the
selective blockage of apoptosis. Administration of Z-VAD in vivo
inhibited caspase 3-like activity (DEVD-MCA) but not caspase 1
activity (YVAD-MCA) and protected effectively against Fas-mediated apoptosis (32). Furthermore, Z-VAD blocked apoptosis
(TGF-b1, staurosporine) but not necrosis (staurosporine) in isolated hepatocytes (41). Kinetic analysis in hepatocyte lysate
showed that Z-VAD is substantially less potent than DEVD-Ald in
inhibiting DEVD-AFC cleavage (33). However, similar inhibition
curves indicated that the mechanism of action was the same; i.e.,
both inhibitors acted as suicide substrates for caspase 3 (33).
YVAD-Ald acted as a suicide substrate for caspase 1 but was a
competitive inhibitor for caspase 3 with a very high Ki of 12.6 mM
(33). Our data agree with these findings. Whereas DEVD-Ald was
a highly potent inhibitor (IC50 of 16.75 nM) for increased caspase
activity in the liver homogenate, YVAD-Ald inhibited this activity
with an IC50 of ;19.6 mM. Because Z-VAD completely prevented
the increase in hepatic caspase activity and apoptosis, it further
supports the conclusion that caspase 3-like proteases are involved
in hepatic parenchymal cell apoptosis in Gal/ET-treated animals.
Z-VAD treatment not only prevented caspase activation and apoptosis but also suppressed liver cell necrosis. Previous studies
showed that parenchymal cell necrosis could be attenuated by antibodies against b2 integrins on neutrophils (35), ICAM-1 on sinusoidal endothelial cells and hepatocytes (14) as well as
VCAM-1 on sinusoidal lining cells (16). Neutrophil transmigration
in hepatic sinusoids has been identified as a critical step for neutrophil-mediated injury in this model (18). Antibodies to ICAM-1
or VCAM-1 blocked neutrophil extravasation and therefore prevented liver cell necrosis (14, 16). These data suggest that neutrophils are essential for the development of hepatocellular necrosis
in this model. Thus, TNF-a initiates two separate responses in the
liver, an apoptotic response and an inflammatory response. Treatment with Z-VAD was started at 3 h after Gal/ET administration.
Since formation of TNF-a (2, 14), neutrophil sequestration in the
hepatic vasculature (18, 35), activation of NF-kB (9, 37), mRNA
formation of ICAM-1 (14, 15), VCAM-1 (16), and selectins (15),
and even in part adhesion molecule protein synthesis (16, 42) already occurs before the 3-h time point, Z-VAD could have not
affected the proinflammatory response in the liver. The only major
events of the inflammatory response that occur after 3 h are neutrophil transmigration and the adherence-dependent cytotoxicity
against hepatocytes. The fact that preventing apoptosis in hepatocytes suppresses neutrophil transmigration suggests that hepatocytes undergoing apoptosis represent a signal for neutrophil migration and attack on parenchymal cells. Generally, apoptosis is
considered a physiologic way to remove unwanted cells without
generating an inflammatory response (43, 44). In contrast, our data
suggest that apoptosis may be able to aggravate an inflammatory
response. How can these opposite viewpoints be reconciled? First
of all, our data do not argue against the fact that single-cell apoptosis will not trigger an inflammatory response. Our data suggest
that in the presence of activated and primed neutrophils in hepatic
sinusoids, a large number of parenchymal cells undergoing apoptosis at the same time can represent a stimulus for these leukocytes to transmigrate and attack. Recently (45), it was reported that
infection of hepatocytes with Listeria monocytogenes causes apoptosis and the generation of a neutrophil chemotactic factor. However, it was unclear whether the increased neutrophil chemotaxis
was actually dependent on apoptosis or was more related to Listeria infection. Nevertheless, this observation is consistent with the
presence of neutrophils around infected hepatocytes in vivo (45).
Neutropenia experiments in this model indicate that a major function of neutrophils is the removal of infected hepatocytes undergoing apoptosis (45). In the Gal/ET model, triggering the transmigration of primed neutrophils may have a similar purpose, and
this may be a general mechanism for removal of a large number of
apoptotic cells. However, as demonstrated after Gal/ET treatment,
the further activation of neutrophils in the liver vasculature bears
the risk of additional damage to healthy tissue. Further studies are
necessary to identify the nature of the chemotactic stimulus generated by apoptotic parenchymal cells.
In summary, a selective activation of caspase 3-like proteases
was observed in the liver that correlated with the development of
apoptosis in parenchymal cells after Gal/ET treatment. Caspase
activation and apoptosis during endotoxemia in vivo was mediated
by TNF-a. Injection of Z-VAD, an effective inhibitor of the
caspase 3 subfamily, prevented caspase activation, apoptosis, and
liver cell necrosis. Z-VAD did not affect neutrophil sequestration
in sinusoids but inhibited transmigration. These data indicate that
activation of the caspase 3 subfamily is critical for the development of parenchymal cell apoptosis. In addition, excessive hepatocellular apoptosis can be an important signal for transmigration
of primed neutrophils sequestered in sinusoids. Thus, proapoptotic
caspases may be a promising therapeutic target for ET- and sepsisrelated liver failure.
3486
26. Leist, M., F. Gantner, S. Jilg, and A. Wendel. 1995. Activation of the 55 kDa
TNF receptor is necessary and sufficient for TNF-induced liver failure, hepatocyte apoptosis and nitric oxide release. J. Immunol. 154:1307.
27. Martin, S. J., and D. R. Green. 1995. Protease activation during apoptosis: death
by a thousand cuts? Cell 82:349.
28. Fraser, A., and G. Evan. 1996. A license to kill. Cell 85:781.
29. Nagata, S. 1997. Apoptosis by death factor. Cell 88:355.
30. Alnemri, E. S., D. J. Livingston, D. W. Nicholson, G. Salvesen,
N. A. Thornberry, W. W. Wong, and J. Yuan. 1996. Human ICE/CED-3 protease
nomenclature. Cell 87:171.
31. Van de Craen, M., P. Vandenabeele, W. Declercq, I. Van den Brande,
G. Van Loo, F. Molemans, P. Schotte, W. Van Criekinge, R. Beyaert, and
W. Fiers. 1997. Characterization of seven murine caspase family members. FEBS
Lett. 403:61.
32. Rodriguez, I., K. Matsuura, C. Ody, S. Nagata, and P. Vassalli. 1996. Systemic
injection of a tripeptide inhibits the intracellular activation of CPP32-like proteases in vivo and fully protects mice against Fas-mediated fulminant liver destruction and death. J. Exp. Med. 184:2067.
33. Inayat-Hussain, S. H., C. Couet, G. M. Cohen, and K. Cain. 1997. Processing/
activation of CPP32-like proteases is involved in transforming growth factor
b1-induced apoptosis in rat hepatocytes. Hepatology 25:1516.
34. Shimizu, S., Y. Eguchi, W. Kamiike, Y. Akao, H. Kosaka, J.-I. Hasegawa,
H. Matsuda, and Y. Tsujimoto. 1996. Involvement of ICE family proteases in
apoptosis induced by reoxygenation of hypoxic hepatocytes. Am. J. Physiol. 271:
G949.
35. Jaeschke, H., A. Farhood, and C. W. Smith. 1991. Neutrophil-induced liver cell
injury in endotoxin shock is a CD11b/CD18-dependent mechanism.
Am. J. Physiol. 261:G1051.
36. Gantner, F., S. Uhlig, and A. Wendel. 1995. Quinine inhibits release of tumor
necrosis factor, apoptosis, necrosis and mortality in a murine model of septic liver
failure. Eur. J. Pharmacol. 294:353.
37. Essani, N. A., M. A. Fisher, and H. Jaeschke. 1997. Inhibition of NF-kB activation by dimethyl sulfoxide correlates with suppression of TNF-a formation,
reduced ICAM-1 gene transcription and protection against endotoxin-induced
liver injury. Shock 7:90.
38. Dunn, C. J., M. M. Hardee, S. F. Fidler, S. K. Shields, and J. G. Chosay. 1996.
The inflammatory potential of IL-2: local induction of a specific chronic granulomatous lesion in mice. J. Leukocyte Biol. 60:27.
39. Rouquet, N., J. C. Pages, T. Molina, P. Briand, and V. Joulin. 1996. ICE inhibitor
YVAD-cmk is a potent therapeutic agent against in vivo liver apoptosis. Curr.
Biol. 6:1192.
40. Rouquet, N., K. Carlier, P. Briand, J. Wiels, and V. Joulin. 1996. Multiple pathways of Fas-induced apoptosis in primary culture of hepatocytes. Biochem. Biophys. Res. Commun. 229:27.
41. Cain, K., S. H. Inayat-Hussain, C. Couet, and G. M. Cohen. 1996. A cleavagesite-directed inhibitor of interleukin-1b-converting enzyme-like proteases inhibits apoptosis in primary culture of rat hepatocytes. Biochem. J. 314:27.
42. Jaeschke, H., N. A. Essani, M. A. Fisher, S. L. Vonderfecht, A. Farhood, and
C. W. Smith. 1996. Release of soluble intercellular adhesion molecule 1 into bile
and serum in murine endotoxin shock. Hepatology 23:530.
43. Abastado, J. P. 1996. Apoptosis: function and regulation of cell death. Res. Immunol. 147:443.
44. Patel, T., and G. J. Gores. 1995. Apoptosis and hepatobiliary disease. Hepatology
21:1725.
45. Rogers, H. W., M. P. Callery, B. Deck, and E. R. Unanue. 1996. Listeria monocytogenes induces apoptosis of infected hepatocytes. J. Immunol. 156:679.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
8. Nathan, C. F., S. Srimal, C. Farber, E. Sanchez, L. Kabbash, A. Asch, J. Gailit,
and S. D. Wright. 1989. Cytokine-induced respiratory burst of human neutrophils: dependence on extracellular matrix proteins and CD11/CD18 integrins.
J. Cell Biol. 109:1341.
9. Essani, N. A., G. M. McGuire, A. M. Manning, and H. Jaeschke. 1996. Endotoxin-induced activation of the nuclear transcription factor kB and expression of
E-selectin messenger RNA in hepatocytes, Kupffer cells, and endothelial cells in
vivo. J. Immunol. 156:2956.
10. Thornton, A. J., R. M. Strieter, I. Lindley, M. Baggiolini, and S. L. Kunkel. 1990.
Cytokine-induced gene expression of a neutrophil chemotactic factor/IL-8 in human hepatocytes. J. Immunol. 144:2609.
11. Colletti, L. M., S. L. Kunkel, A. Walz, M. D. Burdick, R. G. Kunkel, C. A. Wilke,
and R. M. Strieter. 1996. The role of cytokine networks in the local liver injury
following hepatic ischemia/reperfusion in the rat. Hepatology 23:506.
12. Geller, D. A., M. E. de Vera, D. A. Russell, R. A. Shapiro, A. K. Nussler,
R. L. Simmons, and T. R. Billiar. 1995. A central role for the IL-1b in the in vitro
and in vivo regulation of hepatic inducible nitric oxide synthase. J. Immunol.
155:4890.
13. Duval, D. L., D. R. Miller, J. Collier, and R. E. Billings. 1996. Characterization
of hepatic nitric oxide synthase: identification as the cytokine-inducible form
primarily regulated by oxidants. Mol. Pharmacol. 50:277.
14. Essani, N. A., M. A. Fisher, A. Farhood, A. M. Manning, C. W. Smith, and
H. Jaeschke. 1995. Cytokine-induced hepatic intercellular adhesion molecule-1
(ICAM-1) mRNA expression and its role in the pathophysiology of murine endotoxin shock and acute liver failure. Hepatology 21:1632.
15. Essani, N. A., G. M. McGuire, A. M. Manning, and H. Jaeschke. 1995. Differential induction of mRNA for ICAM-1 and selectins in hepatocytes, Kupffer cells
and endothelial cells during endotoxemia. Biochem. Biophys. Res. Commun. 211:
74.
16. Essani, N. A., M. L. Bajt, S. L. Vonderfecht, A. Farhood, and H. Jaeschke. 1997.
Transcriptional activation of vascular cell adhesion molecule-1 (VCAM-1) gene
in vivo and its role in the pathophysiology of neutrophil-induced liver injury in
murine endotoxin shock. J. Immunol. 158:5941.
17. Jaeschke, H., A. Farhood, M. A. Fisher, and C. W. Smith. 1996. Sequestration of
neutrophils in the hepatic vasculature during endotoxemia is independent of b2
integrins and intercellular adhesion molecule-1. Shock 6:345.
18. Chosay, J. G., N. A. Essani, C. J. Dunn, and H. Jaeschke. 1997. Neutrophil
margination and extravasation in sinusoids and venules of the liver during endotoxin-induced injury. Am. J. Physiol. 272:G1195.
19. Tartaglia, L. A., and D. V. Goeddel. 1992. Two TNF receptors. Immunol. Today
13:151.
20. Vandenabeele, P., W. Declercq, R. Beyaert, and W. Fiers. 1995. Two tumor
necrosis factor receptors: structure and function. Trends Cell Biol. 5:392.
21. Tartaglia, L. A., T. M. Ayres, G. H. W. Wong, and D. V. Goeddel. 1993. A novel
domain within the 55 kd TNF receptor signals cell death. Cell 74:845.
22. Leist, M., F. Gantner, I. Bohlinger, G. Tiegs, P. G. Germann, and A. Wendel.
1995. Tumor necrosis factor-induced hepatocyte apoptosis precedes liver failure
in experimental murine shock models. Am. J. Pathol. 146:1220.
23. Haendeler, J., U. K. Messmer, B. Brune, E. Neugebauer, and S. Dimmeler. 1996.
Endotoxic shock leads to apoptosis in vivo and reduces Bcl-2. Shock 6:405.
24. Morikawa, A., T. Sugiyama, Y. Kato, N. Koide, G.-Z. Jiang, K. Takahashi,
Y. Tamada, and T. Yokochi. 1996. Apoptotic cell death in the response of Dgalactosamine-sensitized mice to lipopolysaccharide as an experimental endotoxic shock model. Infect. Immun. 64:734.
25. Tsuji, H., A. Harada, N. Mukaida, Y. Nakanuma, H. Bluethmann, S. Kaneko,
K. Yamakawa, S.-I. Nakamura, K.-I. Kobayashi, and K. Matsushima. 1997. Tumor necrosis factor receptor p55 is essential for intrahepatic granuloma formation
and hepatocellular apoptosis in a murine model of bacterium-induced fulminant
hepatitis. Infect. Immun. 65:1892.
CASPASE ACTIVATION AND ET-INDUCED LIVER INJURY