Intrahepatic STAT-3 activation and acute phase gene
expression predict outcome after CLP sepsis in the rat
KENNETH M. ANDREJKO, JODI CHEN, AND CLIFFORD S. DEUTSCHMAN
Department of Anesthesia, University of Pennsylvania School of Medicine,
Philadelphia, Pennsylvania 19104-4283
cytokines; tumor necrosis factor-a; mortality; multiple organ
dysfunction syndrome; systemic inflammatory response syndrome; a2-macroglobulin
of the hepatic acute phase
response to inflammation, infection, or injury is interleukin-6 (IL-6) (4). After an inflammatory stimulus,
hepatic secretion of several acute phase proteins, including the antiprotease a2-macroglobulin (A2M) in the rat,
is increased. The increase in A2M is transcriptionally
controlled, primarily by IL-6 (4). One IL-6-activated
nuclear protein transcription factor that increases A2M
transcription is signal transducer and activator of
transcription (STAT)-3 (33). Although cytokines other
than IL-6 can activate STAT-3 (20), studies have shown
that STAT-3 is highly reflective of IL-6 activity in the
liver (20).
IL-6 also may be an important modulator of pathological changes in sepsis, the systemic inflammatory response syndrome (SIRS), and the multiple organ dysfunction syndrome (MODS) (7, 8, 12, 13, 16–18, 24, 27).
AN IMPORTANT MEDIATOR
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
Clinical data, however, are inconclusive. One study has
demonstrated a link between serum IL-6 levels and the
development of SIRS/MODS (22), but most investigations in sepsis indicate highly variable IL-6 levels (7,
10, 12, 16, 18, 26). In sepsis/SIRS/MODS, liver dysfunction is prominent (25, 29). There are no studies directly
linking IL-6 with sepsis-induced hepatic dysfunction.
We have recently found that a form of hepatic dysfunction, decreased transcription of key hepatocellular enzymes, is persistent in lethal double-puncture cecal
ligation and puncture (CLP) but not in nonlethal
single-puncture CLP experimental intra-abdominal sepsis (2, 14). Because IL-6 is essential for recovery and
liver regeneration following partial hepatectomy
(11, 32), this cytokine might also be important in the
prevention of or recovery from sepsis-induced hepatic
dysfunction.
The lack of correlation between serum IL-6 levels
and outcome from SIRS/MODS may reflect the fact that
cytokines can affect target cells via three mechanisms:
1) autocrine, affecting the secreting cell; 2) paracrine,
affecting a nearby cell; and 3) endocrine, affecting a
remote cell. In a nonlethal animal model of sepsis, we
have recently shown that tumor necrosis factor-a
(TNF-a)-dependent transcription factor activation and
acute phase gene transcription did not correlate with
serum TNF-a levels. Rather, intrahepatic processes
appeared to be mediated by Kupffer-endothelial cellderived TNF-a (1) and correlated most directly with
activation of a TNF-linked transcription factor, nuclear
factor-kB, and transcription of the TNF-linked acute
phase reactant a1-acid glycoprotein. Thus to determine
the exact role played by IL-6 in sepsis-induced hepatic
dysfunction it would be useful to examine activation of
STAT-3 and transcription of A2M.
In this investigation we examine the role of IL-6 in
single- and double-puncture CLP. We postulate that
following the nonlethal insult of single-puncture CLP,
activation of STAT-3 and transcription of A2M will
persist. In contrast, these processes will decrease following double-puncture CLP (fulminant sepsis), a disorder
characterized by progressive hepatic dysfunction and
death. Furthermore, we hypothesize that intrahepatic
IL-6 abundance will be more predictive of STAT-3
activation and A2M transcription than serum levels.
METHODS
Induction of sepsis. All animal studies conform to National
Institutes of Health standards for the use of laboratory
animals. Overnight-fasted male Sprague-Dawley rats (Charles
River, Boston, MA) weighing between 250 and 270 g were
used in all experiments. The induction of nonfulminant sepsis
was performed under methoxyflurane anesthesia using cecal
0193-1857/98 $5.00 Copyright r 1998 the American Physiological Society
G1423
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
Andrejko, Kenneth M., Jodi Chen, and Clifford S.
Deutschman. Intrahepatic STAT-3 activation and acute
phase gene expression predict outcome after CLP sepsis in
the rat. Am. J. Physiol. 275 (Gastrointest. Liver Physiol. 38):
G1423–G1429, 1998.—Interleukin-6 (IL-6) regulates hepatic
acute phase responses by activating the transcription factor
signal transducer and activator of transcription (STAT)-3.
IL-6 also may modulate septic pathophysiology. We hypothesize that 1) STAT-3 activation and transcription of a2macroglobulin (A2M) correlate with recovery from sepsis and
2) STAT-3 activation and A2M transcription reflect intrahepatic and not serum IL-6. Nonlethal sepsis was induced in
rats by single puncture cecal ligation and puncture (CLP) and
lethal sepsis via double-puncture CLP. STAT-3 activation and
A2M transcription were detected at 3–72 h and intrahepatic
IL-6 at 24–72 h following single-puncture CLP. All were
detected only at 3–16 h following double-puncture CLP and
at lower levels than following single-puncture CLP. Loss of
serum and intrahepatic IL-6 activity after double-puncture
CLP correlated with mortality. Neither intrahepatic nor
serum IL-6 levels correlated with intrahepatic IL-6 activity.
STAT-3 activation following single-puncture CLP inversely
correlated with altered transcription of gluconeogenic, ketogenic, and ureagenic genes. IL-6 may have both beneficial and
detrimental effects in sepsis. Fulminant sepsis may decrease
the ability of hepatocytes to respond to IL-6.
G1424
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
prepared using 10 µg each of plasmids containing A2M and
ATP synthase (positive control) cDNAs. Plasmids without
inserts served as negative controls. Two hundred fifty microliters of 43 reaction mix (100 mM HEPES, pH 7.4, 10 mM
MgCl2, 10 mM dithiothreitol, 300 mM KCl, and 20% glycerol)
were combined with 1 mCi [32P]UTP, 125 µl 83 triphosphate
mixture (2.8 mM each ATP, GTP, and CTP; Boehringer
Mannheim, Indianapolis, IN) and diluted to a volume of 500
µl to yield the 23 reaction mixture. For each reaction, 100 µl
of thawed nuclei were added to an equal volume of 23
reaction mixture and incubated at room temperature for 30
min. The reaction was stopped by adding 2 µl of DNase 1
(10,000 U/ml in 50% glycerol; Boehringer Mannheim) and
incubated at 37°C for 30 min. Stop buffer (600 µl; 2% SDS,
7 M urea, 350 mM LiCl, 1 mM EDTA, and 10 mM Tris, pH
8.0), 1,000 µl/ml of proteinase K, and 100 µl of tRNA were
added and the mixture was incubated at 40°C for 1 h. The
mixture was phenol-chloroform extracted, trichloroacetic acid
precipitated, and ethanol washed. The pellet was resuspended in 100 µl TES (10 mM Tris, 1 mM EDTA, and 0.5%
SDS), added to 5 ml of hybridization solution (see above), and
hybridized with the targets at 42°C for 48 h. Targets were
washed with 23 saline sodium citrate (SSC; 0.03 M sodium
citrate, pH 7.3, and 0.3 M NaCl) at 65°C for 1 h and then
incubated with 10 mg/ml RNase A in 83 SSC at 37°C for 30
min. Targets were then washed with 23 SSC at 37°C for
1 h. Targets were autoradiographed for ,5 days at 280°C,
and quantitative laser densitometry (Molecular Dynamics)
was performed. Radiographic density was normalized to the
density of ATP synthase. The normalized density at time 0
was arbitrarily set at unity, and means 6 SD were determined.
Determination of IL-6 levels in vena caval blood. Determinations of IL-6 serum concentrations were performed using
an ELISA kit specific for rat IL-6 (BioSource, Camarillo, CA).
Plasma samples were assayed in duplicate.
Immunohistochemistry of IL-6 on rat liver sections. Tissue
fixation and preparation were performed as previously described (1). The primary incubation was for 2 h at room
temperature using a goat anti-rat polyclonal antibody for IL-6
(Santa Cruz Biotechnology). The secondary incubation, also
at room temperature, used a biotinylated anti-goat antibody
(Vector Laboratories, Burlingame, CA) for 1 h. Chromogenic
staining was performed with metal-enhanced diaminobenzidine substrate (Pierce, Rockford, IL) using immunoperoxidase methodology (New England Nuclear, Wilmington, DE).
Statistics. Statistical significance (P , 0.05) was determined using ANOVA for repeated measures with the Bonferroni test of post hoc significance.
RESULTS
Outcome. The clinical response differed among sham
operation, single-puncture CLP, and double-puncture
CLP. Sham-operated animals recovered from surgery
uneventfully. After single-puncture CLP, rats developed mild signs of sepsis as previously described (2, 14).
These included lethargy, decreases in spontaneous
movement, and poor grooming that resolved by 48 h. In
contrast, 8–10 h following double-puncture CLP, animals became severely ill (14). Water intake decreased,
there was no spontaneous movement observed, eyes
became encrusted, and grooming was absent. By 16 h
animals were extremely lethargic and had diarrhea,
piloerection, and tachypnea. These signs are characteristic of severe late sepsis.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
ligation with a single, 18-gauge puncture (CLP) as previously
described (14). Fulminant sepsis similarly was induced by
ligating the cecum and then puncturing it twice (3, 31).
Sham-operated controls were anesthetized and underwent
laparotomy with cecal manipulation but without cecal ligation or puncture. After surgery, animals were fluid resuscitated with 40 ml/kg of subcutaneously administered sterile
saline and were given free access to water but not food. At 0, 3,
6, 16, 24, 48, and 72 h following single-puncture CLP or sham
operation and at 0, 3, 6, 16, 24, and 48 h after double-puncture
CLP, animals were reanesthetized with a 50 mg/kg intraperitoneal injection of pentobarbital. Vena caval blood was obtained, and liver tissue was either harvested for nuclear
protein or nuclei, or the entire liver was perfusion fixed with
2% paraformaldehyde.
Isolation and preparation of nuclear extracts. Isolation of
nuclear proteins was performed as previously described (14,
15). All procedures were performed at 4°C, and all buffers
contained proteinase and phosphatase inhibitors. Briefly,
liver tissue was homogenized, nuclei were separated by
ultracentrifugation, the nuclear pellet was lysed, and protein
was isolated. Protein concentrations were determined using
the Bradford method.
Tissue fixation. Livers were perfused via the portal vein
with PBS (pH 7.3) for 5 min to remove red cells. This was
followed by perfusion for 5 min with 2% paraformaldehyde in
PBS (pH 7.3) at a flow rate of 20 ml/min. The perfused liver
was excised, cut into slices, and fixed in the paraformaldehyde solution for 2 h at 4°C. Slices were paraffin embedded
and cut into 7-µm sections. Sections were adhered to poly-Llysine-coated glass slides and dried overnight at 37°C.
Electrophoretic mobility shift assay and supershift analysis. Hepatic nuclear extraction and binding reactions were
performed as previously described with minor modifications
(11, 14, 15). Briefly, a preannealed HPLC-purified oligonucleotide from the serum-inducible-factor binding element in the
c-fos promoter was end-labeled with [g-32P]ATP and used as a
probe for STAT-3 binding activity. An excess of probe was
incubated with 2.5 µg of nuclear extract plus 1 µg polydeoxyinosinic-deoxycytidylic acid, a nonspecific DNA competitor,
for 15 min at room temperature. Other samples were supershifted by the addition of 3 µl of STAT-3 antibody (Santa Cruz
Biotechnology, Santa Cruz, CA) for 1 h after the addition of
labeled oligonucleotide. In cold competition experiments,
unlabeled oligonucleotide was incubated with nuclear extracts for 15 min before the addition of radiolabeled probe.
Each sample was subjected to electrophoresis on a 5% nondenaturing polyacrylamide gel. Gels were then dried, and
autoradiography was performed. Radiographic density of gels
containing samples following sham operation, single-puncture CLP, and double-puncture CLP was determined via
densitometry. Values at time 0 were arbitrarily set at unity,
and other results were normalized to this value.
Transcript elongation analysis (nuclear run-on). Nuclei
were isolated, and nuclear run-on was performed using a
modification of a previously reported method (2). All procedures were performed at 4°C. Briefly, liver tissue was harvested, minced, homogenized in 10 mM Tris · HCl, pH 8.0, 0.3
M sucrose, 3 mM CaCl2, 2 mM magnesium acetate, 0.1 mM
EDTA, and 0.1 mM phenylmethylsulfonyl fluoride, filtered,
and layered onto a cushion buffer containing 50 mM Tris · HCl,
pH 8.0, 2 M sucrose, 5 mM magnesium acetate, 0.1 mM
EDTA, and 0.1 mM phenylmethylsulfonyl fluoride. Ultracentrifugation was performed at 100,000 g for 30 min. The pellet
was resuspended in 25% glycerol, 5 mM magnesium acetate,
0.1 mM EDTA, and 5 mM dithiothreitol, and stored at
2120°C until use. Nitrocellulose-immobilized targets were
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
which is believed to correlate directly with intrahepatic
IL-6 activity (11), was measured following sham operation, single-puncture CLP, and double-puncture CLP
(Fig. 1). No STAT-3 DNA binding activity could be
detected in time 0 samples. Low levels of STAT-3
binding (,10% of that detected following singlepuncture CLP and ,20% of that following doublepuncture CLP) were detected 3 h following sham
operation. Maximal STAT-3 DNA binding activity was
detected 3 h after single-puncture CLP and doublepuncture CLP but levels following single-puncture CLP
were nearly twice those observed following doublepuncture CLP. At all other time points the differences
were equally more pronounced. At 6 and 16 h following
double-puncture CLP, levels of STAT-3 activity reached
a plateau, whereas levels following single-puncture
CLP continued to decline. However, levels were 2.5-fold
Fig. 1. A: representative electrophoretic mobility shift assay (EMSA) detailing STAT-3 DNA binding activity
following sham operation, single-puncture cecal ligation and puncture (CLP), and double-puncture CLP. DNA
binding activity was determined by EMSA. T and subscript indicate time (in h) following sham operation,
single-puncture CLP, or double-puncture CLP. In cold competition (second panel from left) either 10-fold (10 X) or
100-fold (100 X) excess of unlabeled oligonucleotide was added. In ‘‘supershift’’ analysis (third panel from left) an
antibody specific for STAT-3 was added. B: quantification of relative activity of STAT-3 following sham operation,
single-puncture CLP, and double-puncture CLP. Each data point represents mean 6 SD of values from 3 different
animals. Mean value at time 0 arbitrarily set equal to unity; values at other time points normalized to this. Time
following intervention is shown on x-axis. * Significantly different from time 0. † Significantly different from sham
operation value at same time point. # Double-puncture CLP value significantly different from single-puncture CLP
value at same time point. DNA binding activity was determined using EMSA.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
There were no deaths following sham operation or
single-puncture CLP but a significant number of animals died following double-puncture CLP. Six animals
were killed following an overnight fast, 36 animals
underwent sham operation, 36 underwent singlepuncture CLP, and 40 underwent double-puncture CLP.
The intention was to kill six animals in each group
at each time point (3, 6, 16, 24, 48, and 72 h). No
double-puncture animals died before 16 h, but mortality was 50% at 24 h, only three animals survived to 48
h, and none survived to 72 h. Overall mortality in the
double-puncture CLP group was 70%. Four extra animals were needed in the double-puncture CLP group so
that three animals could be studied at the 48-h time
point.
STAT-3 activation in liver nuclear extracts via electrophoretic mobility shift assay. Activation of STAT-3,
G1425
G1426
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
Fig. 2. A: representative transcript elongation assay of
a2-macroglobulin (A2M) and ATP synthase (ATPS) transcription following sham operation, single-puncture CLP,
and double-puncture CLP. B: quantification of relative
rates of transcription of A2M. Radiographic density
from individual elongation analysis divided by ATP
synthase density from same animal. Each data point
represents mean 6 SD of values from 3 different
animals. Mean value at time 0 arbitrarily set equal to
unity; values at other time points normalized to this.
Time following intervention is shown on x-axis. * Significantly different from time 0. † Significantly different
from sham operation value at same time point. # Doublepuncture CLP value significantly different from singlepuncture CLP value at same time point.
patic TNF-a abundance correlated with activation of a
TNF-dependent transcription factor and expression of
a TNF-dependent acute phase reactant (1), we used
immunohistochemistry to see whether a similar correlation existed among intrahepatic IL-6 abundance,
STAT-3 activation, and A2M transcription (Fig. 3). No
IL-6 was detected in the liver at any time point after
sham operation. Intrahepatic IL-6 was detected in
Kupffer and endothelial cells (Fig. 3, black arrows) at
24, 48, and 72 h following single-puncture CLP. More
IL-6 staining was detected at 48 and 72 h than at 24 h
after single-puncture CLP. After double-puncture CLP,
IL-6 was first detected in liver samples 6 and 16 h after
the procedure. There was no IL-6 staining at later time
points. We detected no correlation between IL-6 abundance and the magnitude of either STAT-3 activation or
A2M transcription in the liver after single-puncture
CLP. However, the loss of detectable IL-6 in the liver
following double-puncture CLP correlated with the loss
of STAT-3 binding activity and A2M transcription
IL-6 serum concentrations. Because previous study
after single-puncture CLP has indicated that serum
levels of TNF-a do not correlate with intrahepatic
levels or activity, we examined IL-6 serum concentrations in vena caval blood (Fig. 4). Low levels were
detectable at time 0 and following sham operation.
Serum levels after single-puncture CLP were statistically elevated over time 0 and over sham operation only
at 16 h. In contrast, peak serum levels following
double-puncture CLP were statistically increased 6 and
16 h after the insult. Peak levels after double-puncture
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
greater in single-puncture CLP than in double-puncture CLP at these time points. No STAT-3 binding
activity was detected 24 and 48 h after double-puncture
CLP. These data demonstrate that IL-6 activity was
more pronounced following single-puncture CLP than
following double-puncture CLP, but a sustained elevation was noted following double-puncture CLP.
A2M transcription in rat liver. Transcription of A2M
also provides an indication of intrahepatic IL-6 activity
(4). Nuclear run-on was used to determine timedependent changes in A2M transcription following
sham operation, single-puncture CLP, and doublepuncture CLP (Fig. 2). After sham operation, we detected a twofold elevation of A2M transcription at 6 h.
A2M transcription increased fivefold 3 h after both
single- and double-puncture CLP. This mild increase
was sustained up to 16 h following double-puncture
CLP. Transcription of A2M rose even further following
single-puncture CLP, with a maximal 16-fold increase
observed at 24 h. Levels then declined. This differed
from double-puncture CLP, in which no A2M transcription could be detected at 16, 24, and 48 h. As in previous
studies, transcription of ATP synthase was unchanged
after single- or double-puncture CLP (1, 14). Although
there was no correlation between maximal STAT-3
binding activity and A2M transcription following singlepuncture CLP, we detected a temporal correlation
between the loss of both STAT-3 binding activity and
A2M transcription 24 h after double-puncture CLP.
Immunohistochemical detection of IL-6 in liver tissue. Because we have previously shown that intrahe-
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
G1427
CLP were eight times greater than time 0, sham
operation, and single-puncture CLP. Levels decreased
to baseline 24 h after double-puncture CLP and were
statistically lower than values for sham operation and
single-puncture CLP at 48 h. Serum IL-6 levels following single-puncture CLP did not correlate with STAT-3
activity, A2M transcription, or intrahepatic IL-6 abundance. Although no correlation was evident at 3, 6, and
16 h, the abrupt decrease in serum IL-6 levels 24 h after
double-puncture CLP correlated with the loss of STAT-3
binding activity, A2M transcription, and intrahepatic
IL-6 abundance.
Fig. 4. Serum IL-6 levels following sham operation, single-puncture
CLP, and double-puncture CLP. Blood samples were obtained from
inferior vena caval puncture before rat was killed; levels determined
by ELISA. * Significantly different from time 0. † Significantly different from sham operation value at same time point. # Double-puncture
CLP value significantly different from single-puncture CLP value at
same time point.
DISCUSSION
In this study we demonstrate that loss of STAT-3
activity, transcription of A2M, intrahepatic IL-6 abundance, and immunoreactive IL-6 in the serum correlate
with mortality in sepsis. We further show that neither
serum IL-6 levels nor intrahepatic IL-6 abundance
correlate in magnitude or time with two reliable indicators of intrahepatic IL-6 activity, STAT-3 activity, or
A2M transcription (4, 11), except in the terminal phase
of fulminant sepsis.
This study focuses on the role of IL-6 in hepatic
dysfunction during sepsis/SIRS/MODS. Regel et al. (25)
have found that the liver is the second most commonly
affected organ in MODS, surpassed only by the lung.
We have previously characterized early hepatic dysfunction by examining the transcription of a series of
liver-specific genes in animal models of sepsis (2, 14).
These genes are involved in essential metabolic processes such as gluconeogenesis, ketogenesis, and ureagenesis. In mild sepsis (single-puncture CLP), transcription of these genes decreases and then recovers as
the insult resolves. However, transcription of the gluconeogenic enzyme glucose 6-phosphatase does not return to normal in the face of a fulminant, highly lethal
septic insult, double-puncture CLP (14). Thus singlepuncture CLP is associated with mild, reversible hepatic dysfunction, whereas the alteration associated
with double-puncture CLP is irreversible.
Our data indicate that IL-6 is important in mediating
two distinct, opposing effects in the liver in sepsis. After
single-puncture CLP, IL-6 activity, as measured by
STAT-3 activation and A2M transcription, inversely
correlates with the sepsis-induced decrease in the
transcription of hepatic gluconeogenic, ketogenic, and
ureagenic genes. Specifically, the increase in STAT-3
activity and A2M transcription parallel, in magnitude
and in time, the decrease in transcription of phospho-
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
Fig. 3. Intrahepatic interleukin-6 (IL-6) abundance following sham operation, single-puncture CLP, and doublepuncture CLP. Numbers after abbreviations indicate time (in h) following sham operation (S), single-puncture CLP
(C), and double-puncture CLP (2C). Sections obtained from liver fixed using paraformaldehyde perfusion. Primary
treatment was with a polyclonal goat antibody to rat IL-6, and secondary (detection) staining was with biotinylated
anti-goat IgG. Examples of Kupffer or endothelial cells with positive IL-6 staining indicated by arrows.
G1428
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
TNF and IL-1 are known to enhance IL-6-induced A2M
transcription and activate transcription factors other
than STAT-3 (4, 19, 23). Intrahepatic TNF-IL-1 activity
peaks later than IL-6 activity (1, 15). The enhancement
afforded by this increase could explain increased A2M
transcription despite declining STAT-3 activity. Alternatively, Wen et al. (30) have shown that both serine and
tyrosine phosphorylation are required for maximal
STAT-3 transcriptional activity (30). Thus there may be
a difference between maximal DNA binding activity
and maximal transcriptional activation.
Another important ramification of this study involves the role of serum levels of IL-6 (and other
cytokines) in sepsis, SIRS, or MODS. In septic shock,
death results from cardiovascular collapse secondary to
vasodilatation. During clinical septic shock or following
direct experimental insults, serum TNF, IL-1, and IL-6
levels may correlate with pathophysiological changes.
However, the mechanism of death in patients with
SIRS/MODS is not shock but progressive dysfunction
in multiple organ systems (5, 13). We have previously
shown that TNF-dependent processes correlate with
intrahepatic TNF-a abundance and not with serum
TNF-a levels (1). Neither serum nor intrahepatic IL-6
correlates with two markers of IL-6 activity in sepsis.
This emphasizes that strategies designed to eliminate
or block circulating cytokines are unlikely to be successful.
Perhaps most importantly of all, these data highlight
the complex biology of cytokines such as IL-6 in sepsis.
Levels of IL-6, TNF-a, and IL-1 in the circulation
during clinical sepsis are highly variable. Clinical trials
designed to neutralize circulating TNF-a or IL-1 have
not altered the outcome from SIRS/MODS (10). We
have demonstrated that intrahepatic IL-6 activity is
associated with both detrimental (decreased transcription of gluconeogenic, ketogenic, and ureagenic genes)
and beneficial (improved outcome) effects during sepsis
induced by CLP. In another cell population in a different setting, similarly divergent effects have been shown
for TNF-a (6, 28). In view of this a better understanding
of the complex nature of cytokine-mediated responses
will be essential to develop successful therapeutic
approaches to sepsis.
Drs. Linda Greenbaum and Rebecca Taub critically reviewed the
manuscript, whereas Drs. Antonio DeMaio (Baltimore, MD) and
Mark Clemens (Charlotte, NC) provided the senior author with
ongoing instruction and advice. Finally, the authors gratefully acknowledge Drs. David E. Longnecker and Bryan E. Marshall (Dept.
of Anesthesia, Univ. of Pennsylvania) for advice, support, and a
unique environment of academic opportunity.
This work was supported in part by National Institute of Diabetes
and Digestive and Kidney Diseases Grant K08-DK-028179 (C. S.
Deutschman).
STAT-3 consensus oligonucleotide was provided by Dr. Rebecca
Taub (University of Pennsylvania, Philadelphia, PA). The probe for
A2M was obtained from the American Type Culture Collection
(Rockville, MD).
Address for reprint requests: C. S. Deutschman, Dept. of Anesthesia, Dulles 773/Hospital of the Univ. of Pennsylvania, 3400 Spruce
St., Philadelphia, PA.
Received 12 April 1998; accepted in final form 13 August 1998.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
enolpyruvate carboxykinase, glucose 6-phosphatase,
carnitine palmitoyltransferase, acetyl-CoA acyltransferase, and ornithine transcarbamylase seen after singlepuncture CLP (2). Studies in cultured cells and nonseptic animals have demonstrated that IL-6 reduces
transcription of phosphoenolpyruvate carboxykinase
and glucose 6-phosphatase (9, 21). However, because a
STAT binding domain has not been identified in the
promoters of any of these genes, it is unlikely that
STAT-3 directly mediates the decrease in transcription.
IL-6, via STAT-3 activation, might also mediate hepatic dysfunction by activating a pathway that culminates in cell death, either via necrosis or apoptosis. It is
well known that the double puncture model of CLP is
associated with the late appearance of hepatic transaminases in the blood, indicative of necrosis (3, 31).
Transaminases are not elevated following singlepuncture CLP. We have examined apoptosis following
both forms of CLP and found that it is indeed increased
(L. Bellin, K. M. Andrejko, J. Chen, and C. S. Deutschman, unpublished data).
The correlation between decreased transcription of
gluconeogenic, ketogenic, and ureagenic genes and
increased IL-6 activity supports the hypothesis that the
effects of IL-6 in sepsis are detrimental. However, a
second line of evidence indicates that IL-6 activity is
beneficial. Between 16 and 24 h following doublepuncture CLP, there is a loss of STAT-3 activation, A2M
transcription, detectable IL-6 abundance in the liver,
and immunoreactive IL-6 in the blood. At this point in
time, outcome following single and double-puncture
CLP diverges. In previous work, we found that doublepuncture CLP persistently decreases the transcription
of glucose 6-phosphatase (14). The effect of singlepuncture CLP is not persistent, and the difference
between the two forms of sepsis becomes apparent 16 h
after the original insult (14). Therefore, IL-6 may
function as a ‘‘recovery factor’’ from sepsis and modulate the reversal of hepatic dysfunction. Although the
mechanism is unknown, IL-6 is essential in hepatic
regeneration (11, 32). A similar effect could contribute
to recovery from sepsis.
Double-puncture CLP seems to attenuate the hepatocellular response to IL-6. At 6 and 16 h following this
fulminant insult, STAT-3 activation and A2M transcription are less than those observed after single-puncture
CLP despite higher levels of both intrahepatic and
circulating IL-6. This may reflect a change in an
intracellular, IL-6-linked pathway or that circulating
cytokines do not have access to hepatocytes. Alternatively, the detection of IL-6 by immunohistochemistry
may have no functional importance, indicating nothing
more than the accumulation of IL-6 in Kupffer and
endothelial cells, or normalization of STAT-3 density to
values at time 0 may be invalid. The abundance of IL-6
in nonparenchymal cells at 6 and 16 h following
double-puncture CLP also could reflect an inability of
these cells to secrete cytokines.
Our data also demonstrate a discrepancy between
the time course of STAT-3 activation and A2M transcription following single-puncture CLP. Cytokines such as
INTRAHEPATIC INTERLEUKIN-6 ACTIVITY IN SEPSIS
REFERENCES
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
hepatectomy endotoxemia and sepsis. Arch. Surg. 129: 1159–
1164, 1994.
Gardlund, B., J. Sjolin, A. Nilsson, M. Roll, C.-J. Wickerts,
and B. Wretlind. Plasma levels of cytokines in primary septic
shock in humans: correlations with disease severity. J. Infect.
Dis. 172: 296–301, 1995.
Gerhartz, C., B. Heesel, J. Sasse, U. Hemmann, C. Landgraf, J. Schneider-Mergener, F. Horn, P. C. Heinrich, and
L. Graeve. Differential activation of acute phase response
factor/STAT3 and STAT1 via the cytoplasmic domain of the
interleukin 6 signal transducer gp130. I. Definition of a novel
phosphotyrosine motif mediating STAT1 activation. J. Biol.
Chem. 271: 12991–12998, 1996.
Horvath, C. M., and J. E. Darnell, Jr. The state of the STATs:
recent developments in the study of signal transduction to the
nucleus. Curr. Opin. Cell Biol. 9: 233–239, 1997.
Metzger, S., N. Goldschmidt, V. Barash, T. Peretz, O. Drize,
J. Shilyansky, E. Shiloni, and T. Chejek-Shaul. Interleukin-6 secretion in mice is associated with reduced glucose
6-phosphatase and liver glycogen levels. Am. J. Physiol. 273
(Endocrinol. Metab. 36): E262–E267, 1997.
Partrick, D. A., F. A. Moore, E. E. Moore, W. L. Biffl, A.
Sauaia, and C. C. J. Barnett. The inflammatory profile of
interleukin-6, interleukin-8 and soluble cell adhesion molecule-1
in post-injury multiple organ failure. Am. J. Surg. 172: 425–431,
1997.
Piekorz, R. P., C. Nemetz, and G. M. Hocke. Members of the
family of IL-6-type cytokines activate Stat5a in various cell
types. Biochem. Biophys. Res. Commun. 236: 438–443, 1997.
Pinsky, M. R. Clinical studies on cytokines in sepsis: role of
serum cytokines in the development of multiple-systems organ
failure. Nephrol. Dial. Transplant. 9 Suppl.: S94–S98, 1994.
Regel, G., M. Grotz, T. Weltner, J. A. Strum, and H.
Tscherne. Patterns of organ failure following severe trauma.
World J. Surg. 20: 422–429, 1996.
Tang, G. H., C. D. Kuo, T. C. Yen, H. S. Kuo, K. H. Chan, H. W.
Yien, and T. Y. Lee. Perioperative plasma concentrations of
tumor necrosis factor-alpha and interleukin-6 in infected patients. Crit. Care Med. 24: 423–428, 1996.
Walley, K. R., N. W. Lukas, T. J. Standiford, R. M. Strieter,
and S. L. Kunkel. Balance of inflammatory cytokines related to
the severity and mortality of murine sepsis. Infect. Immun. 64:
4733–4738, 1996.
Wang, C.-Y., M. W. Mayo, and A. S. Baldwin, Jr. TNF and
cancer therapy-induced apoptosis: potentiation by inhibition of
NF-kB. Science 274: 784–787, 1996.
Wang, P., and I. H. Chaudry. Mechanism of hepatocellular
dysfunction during hyperdynamic sepsis. Am. J. Physiol. 270
(Regulatory Integrative Comp. Physiol. 39): R927–R938, 1996.
Wen, Z., Z. Zhong, and J. E. Darnell, Jr. Maximal activation of
transcription by Stat1 and Stat3 requires both tyrosine and
serine phosphorylation. Cell 82: 241–250, 1995.
Wichterman, K. A., A. E. Baue, and I. H. Chaudry. Sepsis and
septic shock: a review of laboratory models and a proposal. J.
Surg. Res. 29: 189–210, 1980.
Yamada, Y., I. Kirillova, J. J. Peschon, and N. Fausto.
Initiation of liver growth by tumor necrosis factor: deficient liver
regeneration in mice lacking type 1 tumor necrosis factor receptor. Proc. Natl. Acad. Sci. USA 94: 1441–1446, 1997.
Zhang, D., M. Sun, D. Samols, and I. Kushner. STAT3
participates in transcriptional activation of c-reactive protein by
interleukin-6. J. Biol. Chem. 271: 9503–9509, 1996.
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.1 on July 28, 2017
1. Andrejko, K. M., and C. S. Deutschman. Acute phase gene
expression correlates with intrahepatic tumor necrosis factor-a
abundance but not plasma TNF levels during sepsis/SIRS in the
rat. Crit. Care Med. 24: 1947–1952, 1996.
2. Andrejko, K. M., and C. S. Deutschman. Altered hepatic gene
expression in fecal peritonitis: changes in transcription of gluconeogenic, b-oxidative and ureagenic genes. Shock 7: 164–167,
1997.
3. Baker, C. C., I. H. Chaudry, H. O. Gaines, and A. E. Baue.
Evaluation of factors affecting mortality rate after sepsis in a
murine cecal ligation and puncture model. Surgery 94: 331–335,
1983.
4. Baumann, H., and J. Gauldie. The acute phase response.
Immunol. Today 15: 74–80, 1994.
5. Beal, A. L., and F. B. Cerra. Multiple organ failure syndrome in
the 1990s: systemic inflammatory response and organ dysfunction. JAMA 271: 226–233, 1994.
6. Beg, A. A., and D. Baltimore. An essential role for NF-kB in
preventing TNF-a induced cell death. Science 274: 782–784,
1996.
7. Blackwell, T. S., and J. W. Christman. Sepsis and cytokines:
current status. Br. J. Anaesth. 77: 110–117, 1996.
8. Casey, L. C., R. A. Balk, and R. C. Bone. Plasma cytokine and
endotoxin levels correlate with survival in patients with sepsis
syndrome. Ann. Intern. Med. 119: 771–778, 1993.
9. Christ, B., A. Nath, P. C. Heinrich, and K. Jungermann.
Inhibition by recombinant interleukin-6 of the glucagon-dependent induction of phosphoenolpyruvate carboxykinase and the
insulin-dependent induction of glucokinase gene expression in
cultured rat hepatocytes: regulation of gene transcription and
messenger RNA degradation. Hepatology 20: 1577–1583, 1994.
10. Christman, J. W., E. P. Holden, and T. S. Blackwell. Strategies for blocking the systemic effects of cytokines in the sepsis
syndrome. Crit. Care Med. 23: 955–963, 1995.
11. Cressman, D. E., L. E. Greenbaum, R. A. DeAngelis, G.
Ciliberto, E. E. Furth, and R. Taub. Liver failure and defective
hepatocyte regeneration in interleukin-6 deficient mice. Science
274: 1379–1383, 1996.
12. Damas, P., D. Ledoux, M. Nys, Y. Vrindts, D. De Groote, P.
Franchimont, and M. Lamy. Cytokine serum levels during
severe sepsis in humans: IL-6 as a marker of severity. Ann. Surg.
215: 356–362, 1992.
13. Deitch, E. A. Multiple organ failure: pathophysiology and
potential future therapy. Ann. Surg. 216: 117–134, 1992.
14. Deutschman, C. S., K. M. Andrejko, B. A. Haber, E. Elenko,
R. Harrison, and R. Taub. Sepsis-induced depression of rat
glucose-6-phosphatase gene expression and activity. Am. J.
Physiol. 273 (Regulatory Integrative Comp. Physiol. 42): R1709–
R1718, 1997.
15. Deutschman, C. S., B. A. Haber, K. Andrejko, D. E. Cressman, R. Harrison, E. Elenko, and R. Taub. Increased expression of cytokine induced neutrophil chemoattractant (CINC) in
septic rat liver. Am. J. Physiol. 271 (Regulatory Integrative
Comp. Physiol. 40): R593–R600, 1996.
16. Dofferhoff, A. S., V. J. J. Bom, H. G. de Vries-Hospers, J. van
Inguen, J. van der Meer, B. P. C. Hazenberg, P. O. M.
Mulder, and J. Weits. Patterns of cytokines, plasma endotoxin,
plasminogen activator inhibitor and acute phase proteins in
severe sepsis. Crit. Care Med. 20: 185–192, 1992.
17. Enayati, P., M. F. Brennan, and Y. Fong. Systemic and liver
cytokine activation: implications for liver regeneration and post-
G1429