| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Department of Internal Medicine, Section of Gastroenterology and Endocrinology, University of Göttingen, Göttingen, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
reduced N-CAM-coding mRNAs, whereas induction of
I-CAM-1- and V-CAM-1-specific transcripts increased severalfold.
In situ, messengers specific for I-CAM-1 and
V-CAM-1 were induced 3 hours after CCl4 treatment (thereby
preceding mononuclear cell infiltration starting at 12 hours),
were expressed at maximal levels 9–12 hours after CCl4
application, and decreased afterwards. I-CAM-1 and V-CAM-1
immunoreactivity increased in a linear fashion starting 3 hours after
CCl4-induced liver injury, was detected in highest
amounts at 24–48 hours characterized by maximal cell
infiltration, and returned to baseline values at 96 hours.
Interestingly, the induction/repression of CAM-specific
messengers paralleled the time kinetics of tumor necrosis
factor-/transforming growth factor-ß1 expression in injured liver.
HSC might be important during the onset of hepatic tissue injury as
proinflammatory elements and might interact with I-CAM-1 and V-CAM-1
ligand-bearing cells, namely lymphocyte function-associated
antigen-1- or Mac-1/very late activation antigen-4-positive
inflammatory cells, thereby modulating the recruitment and
migration of mononuclear cells within the perisinusoidal space of
diseased livers.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hepatic stellate cells (HSC), also designated as Ito cells, fat storing cells, or lipocytes, play a major role in vitamin A metabolism and are presently regarded as the principal cell type responsible for matrix accumulation during liver repair reactions including fibrosis.6-8 HSC are situated in the space of Disse between the sinusoidal endothelium and hepatocytes and exhibit long cytoplasmic processes that underlie the endothelium and embrace the sinusoid but also have contact with hepatocytes. Because HSC reflect the cell population immediately contacting transmigrated leukocytes on one side and hepatocytes on the other side, they could play a role in the recruitment of inflammatory cells into the liver parenchyma and subsequently in their targeting to damaged hepatocytes.
To address this question, the present study analyzed HSC with respect to the expression of cell adhesion molecules and their regulation by mediators involved in hepatic tissue repair reactions. Among cell adhesion molecules (CAMs) the report concentrates on intercellular cell adhesion molecule-1 (I-CAM-1) and vascular cell adhesion molecule-1 (V-CAM-1), because the I-CAM/ lymphocyte function associated antigen (LFA)-1 and V-CAM/very late activation antigen (VLA)-4 adhesion pathways have been described as fundamental for leukocyte recruitment at sites of inflammation.1-3,9,10 In addition, the regulation of neural cell adhesion molecule (N-CAM, also termed CD56) was analyzed because N-CAM, known to be expressed by HSC following activation, might be involved in the migration of CD56-positive lymphocytes or HSC and in the termination of HSC proliferation induced by tissue injury.11,12
Basic features of the response of HSC to hepatic tissue injury are phenotypical and functional changes, a process called activation.6-8 Activation includes HSC proliferation, transforming star-shaped cells rich in vitamin A to vitamin A-deficient cells of myofibroblast-like appearance (activated HSC) displaying contractile properties. Furthermore, activation is characterized by differential gene expression of connective tissue components, matrix-degrading enzymes, and their inhibitors, resulting in matrix accumulation colocalized with activated HSC. Interestingly, this in vivo activation process strongly resembles the morphological and functional changes observed in HSC during primary culture. Therefore HSC in vitro are commonly used as a model to study the role of these cells during hepatic tissue repair, which additionally offers the unique opportunity to study the functional role of these cells at different activation steps reflecting different phases of tissue injury. Because the potential involvement of HSC in leukocyte recruitment might be affected by their differentiation stage, the expression and regulation of CAMs was studied in HSC in vitro at different steps of activation.
To clarify whether the data obtained from former in vitro studies are relevant to in vivo conditions, expression of I-CAM-1 and V-CAM-1 was analyzed in the carbon tetrachloride (CCl4) in vivo model for acute liver damage accompanied by hepatic inflammation. Using this model, the time kinetics and tissue distribution of CAM expression, the infiltration of mononuclear cells, and the expression of cytokines, identified by the in vitro studies as strong inducers or repressors of CAM in HSC, were analyzed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Wistar rats were provided by Charles River (Sulzfeld, Germany) and received humane care in compliance with the institution's guidelines and National Institutes of Health guidelines.
cDNA Probes
To detect transcripts specific for CAM, polymerase chain reaction (PCR) generated cDNAs directed against rat N-CAM,12 rat I-CAM-1, which mapped to positions 529–782 of the published sequence,13 and against rat V-CAM-1, corresponding to positions 352–741 of the published sequence,14 were used.
Transforming growth factor (TGF)-ß1-specific messengers were detected
using a PCR-generated cDNA directed against rat TGF-ß, which mapped
to positions 763-1063 of the published
sequence.15
Furthermore, a PCR-generated cDNA
directed against rat tumor necrosis factor (TNF)-, which mapped to
positions 140–509 of the published sequence,16
was used.
In addition, clone pFH154 coding for human fibronectin17 and a cDNA probe specific for human albumin18 were used. To validate quantitative Northern blot results a clone carrying the rat glycerylaldehyde-3-phosphod dehydropenase (GAPDH) cDNA19 or human ß-actin20 were used. Specificity of PCR products mentioned above was confirmed by digestion using appropriate restriction enzymes and by sequencing of cloned PCR products. PCR products were cloned using the TA cloning kit (Invitrogen, San Diego, CA) and sequenced using the Sequenase version 2.0 kit (United States Biochemical, Cleveland, OH). Sequence comparison was performed by Fasta or BestFit alignment programs of the genetics computer group package (Genetics Computer Group, Madison, WI) using standard parameters.21
Antibodies
Monoclonal antibodies directed against rat I-CAM-1 were obtained from Genzyme (Cambridge, MA) and monoclonal antibodies directed against human V-CAM-1 (clone 51–10C9) from Pharmingen (San Diego, CA). A monoclonal anti N-CAM antibody (clone NCAM-OB11) was obtained from Sigma (Munich, Germany). The mAb against desmin, the antiserum directed against mouse IgGs, and the APAAP complex were from Dako (Copenhagen, Denmark). The mAbs directed against the ED1 and ED2 epitopes were from Biermann (Wiesbaden, Germany). The mAbs against smooth muscle alpha action (SMA) and antibodies directed against glial fibrillary acidic protein were from Sigma. The mAb against vimentin was from Boehringer (Mannheim, Germany)
Mediators
Cytokines were from the following sources and were tested at the
concentrations provided below unless otherwise stated in the legends:
TGF-ß1, human, natural, 1 ng/ml (Sigma); insulin-like growth
factor-1 (IGF-1), human, natural, 100 nMol (kindly provided by Dr.
Märki, Ciba Geigy, Basel, Switzerland); platelet-derived growth
factor (PDGF) (Sigma), human, natural, 10 ng/ml; epidermal growth
factor (EGF), human, recombinant, 2.5 ng/ml (Sigma); hepatocyte growth
factor (HGF), human, recombinant, 10 ng/ml (Sigma); TNF-, human,
recombinant, 100 U/ml (Genzyme); interferon-
(IFN), rat,
recombinant, 100 U/ml (Genzyme).
Isolation and Cultivation of HSC and Other Liver Cells
HSC were isolated from rat liver and kept in primary culture as described previously.12,22-26 As assessed by morphology and by the expression of SMA, GFAP, and N-CAM, HSC were considered fully activated at 7 days of primary culture and later, whereas cells cultured for 2 days were classified as resting HSC or HSC at an early stage of activation and cells cultured for 4 days were judged as transitional cells.12,23,27 Purity of freshly isolated cells and cultured cells was assessed as stated earlier.12,22-26
To verify that the CAM expression levels detected in HSC were not derived from other liver cells, hepatocytes, Kupffer cells, and sinusoidal endothelial cells were isolated from rat liver and cultivated as described previously.12,22,24,25,27-29
Cytokine Treatment of HSC
HSC cultured for 2, 4, or 7 days were incubated for 20 hours in 1
ml serum reduced (0.3% fetal calf serum) culture medium with or
without (controls) the mediators, which were used at the concentrations
stated under "mediators." Furthermore, the effects of increasing
concentrations of TNF- (1 U/ml, 10 U/ml, 100 U/ml and 1000 U/ml),
TGF-ß1 (0.1 ng/ml up to 10.0 ng/ml), or IGF-1 (1 nMol up to 100 nMol)
were tested. At the end of the cytokine exposure, cells were processed
for RNA extraction and Northern blot analysis as stated below.
Experiments analyzing the regulation of CAM expression by mediators at
the standard concentrations and 20-hour incubation periods testing HSC
cultured for 2, 4, or 7 days were performed at least twice using at
least two different HSC isolations.
Induction of Acute Liver Damage
Rats were given carbon tetrachloride (CCl4)/maize oil solution (50%, v/v) orally as described earlier.12,27,28 Control animals were given maize oil only.
Immunohistochemistry and Immunocytochemistry
Cryostat sections (5 µm) of normal or acutely damaged liver specimens and cells were fixed in methanol/acetone (5 minutes/10 seconds at -20°C) as described earlier.12,27-29 Cells or tissue sections were examined by the alkaline phosphatase anti-alkaline phosphatase (APAAP) staining procedure. Cells or tissue sections were incubated with specific primary antibodies for 1 hour at 37°C and were then covered for 1 hour at 37°C with an antiserum directed against mouse IgG followed by incubation with the APAAP complex. Nonspecific staining was controlled for by incubation with mouse Ig instead of the specific primary antibody. To compare the tissue distribution patterns of I-CAM-1, V-CAM-1, N-CAM, vimentin, desmin, GFAP, SMA, and ED1, serially cut tissue sections were prepared and analyzed.
RNA Extraction and Northern Blot Analysis
Cells were lysed with guanidinum isothiocyanate and total RNA was extracted as reported.12,23,26 For Northern blot analysis of the basal CAM expression levels, RNA prepared from four different cell preparations was pooled to overcome individual differences. The RNA (5 µg total) was resolved by agarose gel electrophoresis, transferred to nylon membranes, and hybridized with specific 32P-labeled cDNA probes. Hybridization was performed for 2 hours at 68°C using the QuickHyb Kit (Stratagene, La Jolla, CA). Posthybridization washes were performed 2 times for 15 minutes at room temperature and 1 time for 5–15 minutes at 60°C in 2x SSC containing 0.1% sodium dodecyl sulfate. Nylon filters were washed, dried, and exposed to x-ray films at -80°C.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Using cDNAs generated by RT-PCR, I-CAM-1, V-CAM-1, and (as shown
previously12
) N-CAM, coding transcripts were
detected in HSC in vitro (Figure 1A)
. As reported for rat
I-CAM-130
and human or mouse
V-CAM-1,31,32
single I-CAM-1- and
V-CAM-1-specific messengers of 3.0 kb and 3.2 kb in size were present
in HSC from rat liver, respectively. Hybridization of filters
containing RNA samples derived from hepatocytes, Kupffer cells, and
sinusoidal endothelial cells of rat liver both directly after isolation
and at different stages of primary culture indicated that the CAM
expression of HSC was not derived from contaminating cells (data not
shown).
|
. In contrast, in the same set of experiments
N-CAM was expressed following in vitro activation of HSC as
demonstrated earlier12
(Figure 1A)
.
Using antibodies directed against rat I-CAM-1 and human V-CAM-1, a
specific immunoreactivity was noted in HSC (Figure 1B)
. Already at day
2 of primary culture a weak I-CAM-1- and V-CAM-1-specific staining
reaction, distributed in the cytoplasm with some pronouncement toward
the cell surface membrane, was detected in HSC. At day 7 the
CAM-specific staining reaction was localized predominantly at the cell
surface, particularly in areas of cell-cell contact, and resembled the
staining pattern detected in the case of N-CAM.12
Due to the species specificity of the antibodies used, V-CAM-1-specific
staining reactions were generally weaker compared to I-CAM-1 and might
underestimate the amount of V-CAM-1 actually present in HSC but also
within liver tissue (Figure 2)
. The latter suggestion is supported by
the observation that the immunodetection of I-CAM-1 testing cells or
liver tissue was considerably stronger using an antibody directed
against rat I-CAM-1 compared to an antibody directed against human
I-CAM-1 used in the initial experiments.
|
To study whether HSC in situ also express I-CAM-1 ad
V-CAM-1, serially cut tissue sections were examined by the APAAP
staining method (Figure 2)
. In normal
liver I-CAM-1- and V-CAM-1-specific staining reactions were detected in
the sinusoidal/perisinusoidal area of the hepatic parenchyma. In an
attempt to assign the CAM immunoreactivity more precisely, livers were
examined for the presence of desmin, GFAP, SMA, N-CAM, ED1, and ED2.
Desmin and GFAP are known to be present in resting HSC corresponding to
HSC localized in normal liver,27
whereas SMA and
N-CAM become detectable only in activated HSC present in diseased
livers.12
In addition, desmin and SMA are known
to be expressed in myofibroblasts of the blood vessels in the portal
field,12,27
whereas N-CAM is localized to
neural structures.12
The
sinusoidal/perisinusoidal I-CAM-1 and V-CAM-1 staining reactions showed
some colocalization with desmin/GFAP-positive HSC. However, from these
colocalization studies an I-CAM-1/V-CAM-1 positivity of other
sinusoidal liver cells, like sinusoidal endothelial cells or
ED2-positive Kupffer cells, cannot be excluded.
Regulation of CAM Expression in HSC by Growth Factors
To study the modulation of CAM expression, HSC at different steps
of activation were exposed to growth factors (TGF-ß1, IGF-1, PDGF,
EGF, and HGF), because the latter cytokines have been shown to be
essential during tissue repair reactions of the liver and other organs
and are known to modulate CAM expression in a variety of cell culture
models.33-38
In general, growth factors
(prototype TGF-ß1) down-regulated I-CAM-1, and V-CAM-1
mRNA steady-state levels and stimulated N-CAM expression (Figure 3)
.
|
. HGF also stimulated N-CAM expression
but this effect was present mainly in activated HSC and was lower
compared to the other growth factors (data not shown). I-CAM-1 specific mRNAs were diminished following exposure of HSC at days 2 and 4 after plating to TGF-ß1, IGF-1, EGF, and PDGF. In fully activated HSC (7 days after plating) only IGF-1 caused an I-CAM-1 down-regulation, whereas most of the other growth factors displayed no significant effects on I-CAM-1 expression. Interestingly, the regulation of I-CAM-1 expression by PDGF and HGF was dependent on the HSC activation status. PDGF diminished I-CAM-1 expression in HSC at 2 days after plating, was without major effect in HSC at 4 days of primary culture, and stimulated I-CAM-1 steady state levels in fully activated HSC. On the other hand, HGF resulted in significant down-regulation of I-CAM-1 expression only in fully activated cells, whereas in HSC at days 2 and 4 after plating I-CAM-1 mRNA steady-state levels were not affected by HGF exposure (data not shown).
The regulation of V-CAM-1 expression by growth factors resembled the pattern described for I-CAM-1; however, a major difference was noted in the case of IGF-1. Although most of the growth factors (TGF, PDGF, EGF, and HGF) reduced V-CAM-1 expression levels, IGF-1 induced V-CAM-1-coding mRNAs in HSC at all differentiation stages. With the majority of mediators, down-regulation of V-CAM-1 was most prominent in transitional cells (HSC 4 days after cultivation). In HSC 2 days after plating this effect was detectable only in the case of EGF, and in fully activated cells following incubation with PDGF and EGF.
Regulation of CAM Expression in HSC by TNF- and IFN-
Among the inflammatory cytokines TNF- appeared to be the most
potent modulator of CAM expression and displayed a CAM regulation
pattern almost completely opposite that of the growth factors
analyzed (Figure 4)
. TNF-
reduced N-CAM steady-state levels while I-CAM-1- and V-CAM-1-specific
transcripts were induced severalfold. TNF--mediated effects were
present throughout HSC primary culture. However, due to a low N-CAM
expression level in HSC 2 days after plating, the reduction of N-CAM
coding transcripts by TNF- was not apparent at this stage of primary
culture (Figure 4)
. In general, IFN- treatment resulted in CAM
induction/deinduction kinetics similar to TNF-. In contrast to
TNF-, IFN- stimulated mostly V-CAM-1 expression but stimulated
I-CAM-1-specific messengers only in minor amounts and preferentially in
more advanced stages of HSC activation (Figure 4)
.
|
, and IGF-1
Based on the foregoing experiments, TGF-ß1, TNF-, and IGF-1
were identified as potent modulators of CAM expression in HSC at
different stages of activation. Because the primary aim of the present
study was to clarify whether HSC are involved in the recruitment of
inflammatory cells in the initial phase of liver damage, resting HSC
and HSC at an early stage of activation were analyzed with respect to
the dose dependency of the latter mediators (Figure 5)
. TNF- stimulated I-CAM-1 and
V-CAM-1 expression at concentrations of 1 U/ml and maximal induction
was noted at doses of 100 U/ml (Figure 5)
. Low concentrations (0.01
ng/ml) of TGF-ß1 were also able to diminish I-CAM-1 and V-CAM-1 mRNA
steady-state levels, which were reduced most efficiently by TGF-ß1 at
doses between 1 ng/ml and 10 ng/ml (Figure 5)
. Furthermore, IGF-1
exhibited its effects in a dose-dependent manner. A reduction of
I-CAM-1 steady state mRNA levels was noted at a concentration of 1
nMol, being maximal at 100 nMol, whereas V-CAM-1 induction became
evident at 10 nMol and showed extensive induction at 10–100 nMol
concentrations (Figure 5)
. Control hybridizations using
fibronectin-specific cDNAs confirmed previous data in that TGF-ß1 and
TNF- induced fibronectin-specific messengers (Figure 5)
.23,26
However, compared to the CAM modulation,
expression of fibronectin-coding mRNAs was stimulated by the latter
cytokines, generally at higher concentrations starting at 10 U/ml
TNF- and 0.1 ng/ml TGF-ß1. IGF-1 was ineffective in stimulating
fibronectin expression, whereas in the same set of experiments an
induction of HGF-specific mRNAs was present as reported
previously39
(data not shown).
|
, mediators were active within a
20-hour incubation period. As evidenced in single experiments (data not
shown), growth factors like TGF-ß1 and inflammatory cytokines like
TNF- and IFN- were efficient within 8-hour incubation periods to
modulate CAM expression in the cytokine-specific manner mentioned
above. After 8 hours' incubation TNF- and IFN- stimulated
I-CAM-1 and V-CAM-1 expression, which was diminished by growth factors.
The same was true in the case of N-CAM, which was induced by TGF-ß1
and diminished by TNF- 8 hours after exposure to the mediators. Expression of Cell Adhesion Molecules in Acutely Injured Livers
From the data presented so far it became evident that HSC in
vitro and presumably in vivo express I-CAM-1 and
V-CAM-1 and that inflammatory cytokines (prototype TNF-) induced
I-CAM-1 and V-CAM-1 expression, whereas growth factors (prototype
TGF-ß1) diminished their steady-state mRNA levels. To gain insights
into the potential in vivo relevance of these findings, the
time kinetics and tissue distribution of CAM expression were analyzed
by Northern blotting (Figure 6A)
and
immunohistochemistry (Figure 6, B and C)
during liver inflammation
induced by a single application of CCl4. In
parallel, the steady-state mRNA levels of the important mediators,
namely TGF-ß1 and TNF-, were characterized in this experimental
liver injury model.
|
By Northern blot analysis (Figure 6A)
V-CAM-1 and I-CAM-1, which were
present in amounts close to the detection limit in normal livers, were
severalfold induced in the early phase of the response. Both CAMs were
up-regulated 3 hours after CCl4-induced liver
injury, reached maximal mRNA steady state levels at 9–12 hours, and
progressively diminished afterwards, returning to baseline levels about
72 hours after CCl4 application. Interestingly,
the induction/repression of CAM-specific messengers paralleled the time
kinetics of TNF-/TGF-ß1 expression in injured liver.
TNF--specific messengers 1.6 kb in size41
were
also induced at 3 hours post-CCl4 application,
showed highest levels at 9–12 hours, and were down-regulated later on,
exactly matching the CAM induction kinetics. In contrast, hepatic
expression of TGF-ß1-specific transcripts 2.5 kb in size, which have
been shown to repress CAM expression in HSC in vitro, became
elevated at 9 hours and increased to highest quantities at 24–72
hours, matching the CAM repression profile. In control hybridizations
albumin-specific mRNAs decreased in a time-dependent manner to lowest
levels at 24–48 hours after CCl4 administration,
following the course of necrosis. Due to hepatic regeneration, albumin
expression increased with the onset of the recovery phase at 72 hours
and reached baseline levels of normal livers at 96 hours (data not
shown).
By immunohistochemistry V-CAM-1 and I-CAM-1 staining of the
sinusoidal/perisinusoidal area was increased 3 hours after
CCl4 administration compared to normal livers
(Figure 6B)
. I-CAM-1 and V-CAM-1 immunoreactive material increased in a
linear fashion thereafter, reaching maximal levels at 24 hours
(V-CAM-1) and 48 hours (I-CAM-1) (Figure 6B, 6C)
. V-CAM-1- and
I-CAM-1-positive cells were localized mainly in the necrotic area
(Figure 6B, 6C)
. In livers at 96 hours I-CAM-1 and V-CAM-1 staining was
present in amounts near the baseline level of normal livers, a dramatic
reduction in comparison to livers at 48 hours (Figure 6B)
.
CAM expression in injured livers was studied in parallel with marker
proteins, namely vimentin, desmin, GFAP, smooth muscle -actin, ED1,
and ED2, reacting with all mesenchymal, fibroblastic, and mononuclear
cell populations. Compared to normal livers the number of
vimentin-positive cells reflecting all mesenchymal cell populations
involved in tissue repair were increased at 12 hours, followed by a
dramatic incline to 48 hours and a decline to 96 hours (data not
shown). ED1- and ED2-positive cells corresponding to inflammatory cells
and local macrophages (Kupffer cells), respectively, showed the same
kinetics; a significant increase in their numbers was noted at 12 hours
after CCl4 application followed by a further
expansion, reaching maximal levels at 48 hours (Figure 6C)
. However,
eg, at 24 and 48 hours, ED1-positive cells representing inflammatory
cells were localized mainly within the necrotic area, whereas
ED2-positive cells reflecting Kupffer cells were distributed primarily
around tissue necrosis (Figure 6C)
. Desmin- and GFAP-positive cells
also displayed the time kinetics reported for vimentin-positive cells
and were present at 48 hours in higher amounts in the necrotic area
than in the surrounding tissue (Figure 6C)
. As reported previously,
N-CAM-positive cells also became detectable in injured livers. However,
the absolute numbers were severalfold lower compared to I-CAM-1- and
V-CAM-1-positive cells and N-CAM-positive cells were seen only in minor
amounts, mostly within the necrotic area and occasionally in the
surrounding tissue at 24 hours after CCl4
application (data not shown). SMA-positive cells were detected solely
in the vessel walls from 3–12 hours and again at 96 hours. At 24 hours
and especially at 48 hours, SMA-positive cells were also found in the
necrotic area in numbers severalfold lower than desmin- and
GFAP-positive cells (data not shown). At 96 hours, the numbers of
mononuclear cells (Figure 6B)
and desmin- and GFAP-positive cells (data
not shown) were severalfold diminished, but still slightly above the
baseline level of normal livers.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In addition to chemokines, several adhesion molecules act in concert to facilitate migration of leukocytes from the blood to extravascular sites of inflammation. Presently adhesion molecules are grouped into four major families: integrins, the immunoglobulin superfamily members including I-CAM and V-CAM, selectins, and selectin ligands like muzin.48,49 In contrast to chemokines, CAMs are somehow active at later stages of the inflammatory reaction and mediate direct interaction of infiltrating mononuclear cells with resident cells.
As in other organs it has been demonstrated in the liver that CAM expression by sinusoidal endothelial cells is essential for the transmigration of inflammatory cells and that CAMs present on hepatocytes might be important for the interaction of inflammatory cells with their target cells.29,50-59 Although the mechanisms involved in the transmigration of blood cells through the sinusoidal endothelial lining have been elucidated, the migration of leukocytes within the hepatic tissue that results in accumulation of inflammatory cells in areas of liver damage is poorly characterized. Because HSC reflect the cell population immediately contacting transmigrated leukocytes on one side and hepatocytes on the other side, HSC could play a role in the recruitment of inflammatory cells into the liver parenchyma and subsequently in their targeting to damaged hepatocytes. Therefore the present study analyzed the expression and regulation of CAMs by HSC, in particular the expression of I-CAM-1 and V-CAM-1, known to be essential for the inflammatory reaction.
The data presented in this report illustrate that HSC in vitro expressed I-CAM-1, V-CAM-1, and N-CAM, as evidenced by immunocytochemistry and Northern blot analysis. In the case of N-CAM and I-CAM-1 the results confirmed and extended previous studies;12,60 however, as demonstrated in the case of V-CAM, novel CAMs present on HSC were also identified. It seems of special interest that HSC produce V-CAM-1, which has been detected in the liver only in endothelial cells, Kupffer cells, and bile duct epithelia, especially following lipopolysaccharide-induced liver injury or graft rejection.51,52,61
CAM expression by HSC was studied at various time points of primary culture reflecting different activation stages of HSC. Using the latter approach the data demonstrate that HSC in both the resting and activated states express I-CAM-1 and V-CAM-1, whereas, in accordance with previous studies,12 N-CAM expression became detectable after in vitro activation.
The observation that freshly isolated and early cultured HSC classified as resting cells express significant amounts of I-CAM-1-specific transcripts contradicts a recent study by Hellerbrand and coworkers60 demonstrating that I-CAM-1 mRNAs were expressed in activated (day 15 of culture) but not in resting (freshly isolated) HSC as assessed by Northern blot analysis. However, additional time points of primary culture were not analyzed and I-CAM-1 coding transcripts were present in low amounts in freshly isolated HSC using RT-PCR techniques in the Hellerbrand et al report, thereby providing possible explanations for the different observations.
On the other hand one might argue that the I-CAM-1 and V-CAM-1 expression by resting HSC (freshly isolated and 2–3 days after plating) described in this report might be caused by contaminating cells. The latter possibility was examined by Northern blot analysis of other resident liver cell populations, namely hepatocytes, Kupffer cells, and sinusoidal endothelial cells, both directly after isolation and at various stages of primary culture. Results indicated that the I-CAM-1 and V-CAM-1 expression ascribed to HSC was not due to contaminating cells. In addition, further evidence for I-CAM-1 and V-CAM-1 production by HSC was provided by the detection of I-CAM-1 and V-CAM-1 protein in HSC using immunocytochemistry.
To examine whether HSC in situ express CAMs, I-CAM-1 and V-CAM-1 were localized in normal rat liver by immunohistochemistry, illustrating that V-CAM-1 was present in the sinusoidal/perisinusoidal area. Because these CAMs colocalized only in part with desmin- and GFAP-positive HSC and because the more extended in vitro analysis also indicated that Kupffer cells and sinusoidal endothelial cells expressed I-CAM-1 and V-CAM-1, it seems reasonable to conclude that several nonparenchymal liver cell populations including HSC express I-CAM-1 and V-CAM-1 in normal liver, confirming previous data on I-CAM-1 expression in situ.29,51,61,62
Cell adhesion molecules are basic to tissue inflammation responses and
their expression is known to be regulated by a wide range of cytokines
involved in tissue repair. To gain further insights into the potential
role of HSC in hepatic inflammation via CAMs the effect of mediators
known to play an essential role in liver tissue repair, namely
classical inflammatory cytokines and various growth factors, on CAM
expression by HSC was studied. Interestingly, inflammatory cytokines
(prototype TNF-) induced I-CAM-1 and V-CAM-1 expression, whereas
N-CAM steady-state levels were reduced. In contrast, growth factors
(prototype TGF-ß1) generally caused a completely opposite pattern,
diminishing I-CAM-1 and V-CAM-1 mRNA levels but inducing N-CAM
expression.
Although reports of the induction of I-CAM-1 expression by
TNF-60
and of N-CAM by TGF-ß in
HSC12
have been published recently, the present
study provides a more detailed analysis of the cytokine-regulated CAM
expression in liver cells and demonstrates that HSC must be regarded as
an essential element in hepatic inflammation. In addition, the in
vitro studies suggest different roles for the various mediators
during tissue repair (Figure 7)
.
Inflammatory cytokines might be responsible for the initial phase of
inflammation via I-CAM-1 and V-CAM-1 up-regulation, whereas growth
factors like TGF-ß1 might terminate cell recruitment through
down-regulation of l-CAM-1 and V-CAM-1 (Figure 7)
. This
hypothesis was further supported by the time kinetics of expression of
CAM, TNF-, and TGF-ß during hepatic inflammation in
situ, illustrating that the induction and repression of
CAM-specific messengers paralleled the time kinetics of TNF- and
TGF-ß1 expression in injured liver.
|
As demonstrated in a number of experimental settings29,50,53,59,66,67 and during human liver diseases including, eg, viral infections, acute liver allograft rejection, alcoholic hepatitis, and others,54-56,58,61,68-70 I-CAM-1 and V-CAM-1 are up-regulated in the course of hepatic injury, as evidenced by immunohistochemistry and Northern blot analysis of injured tissue or determination of soluble CAMs in the blood. Although numerous studies have been published, especially on I-CAM-1, detailed analyses of the localization of CAMs in injured livers and their time kinetics from onset through recovery are rare.
The data presented in this paper indicate that both I-CAM-1 and V-CAM-1
were strongly up-regulated in areas of tissue necrosis, that this
up-regulation preceded the infiltration of mononuclear cells, and that
I-CAM-1 and V-CAM-1 were diminished during the recovery phase in
parallel with a reduction of cellular infiltration. These results
provide further support for the assumption that both I-CAM-1 and
V-CAM-1 are essential components of leukocyte recruitment into the
inflamed liver. Moreover, the present study provides experimental
evidence that TNF- and TGF-ß seem to be involved in hepatic
inflammation via regulation of CAMs, because in situ the
induction/deinduction kinetics of CAM expression paralleled those of
the mediators and in vitro both cytokines regulated CAM
expression in a cytokine-specific manner.
The role of TNF- as a key cytokine involved in hepatic tissue
repair, particularly liver inflammation, has been documented
extensively. Acute and chronic liver diseases in humans and in animal
models are accompanied by increased TNF- expression both in the
liver and in peripheral mononuclear
cells.58,67,71-73
In the diseased liver TNF-
is expressed by Kupffer cells, mononuclear cells, or T lymphocytes
present within inflammatory cell infiltrate and intrahepatic bile duct
epithelium74,75
and also by hepatocytes during
hepatitis B and C virus infection.76
The
stimulatory effect of TNF- on CAM expression by liver cells analyzed
in vitro is well documented52,57,60
but only a few reports actually demonstrate that TNF- is related to
enhanced CAM expression during liver injury in
situ.58,59,67,77
In contrast to the latter
publications, the present report demonstrates the distinct time
kinetics characterizing the TNF- and CAM induction/deinduction,
thereby presenting an important insight into the specific mechanisms of
liver inflammation.
In contrast to TNF-, the biological significance of TGF-ß1 in the
regulation of CAM expression during liver inflammation has not been a
major research topic. Although it has been demonstrated that TGF-ß1
is transiently enhanced at later stages of acute liver injury (ie, at
approximately 48 hours after CCl4 administration
in the rat), thereby matching the data presented in this
study,73,78
a direct relationship to CAM
expression and subsequently the regulation of cell infiltration has not
yet been suggested. From the data presented in this paper, namely the
TGF-ß1-mediated down-regulation of CAM expression by HSC and the
expression kinetics of TGF-ß1 and CAMs in situ, it is
tempting to speculate that TGF-ß1, probably in combination with
other growth factors, is responsible for termination of the
inflamma-tory reaction through a reduction of CAM expression (Figure 7)
.
In summary, based on the data presented in this paper it seems
reasonable to suggest that the initial phase of liver inflammation
might be triggered through mediators liberated from injured
hepatocytes, which, as shown in the case of IGF-1, stimulate V-CAM-1
expression by HSC (Figure 7)
. In the next step activated resident liver
macrophages (Kupffer cells) and inflammatory cells might augment
I-CAM-1 and V-CAM-1 expression by HSC through secretion of TNF- and
IFN-, thereby leading to further recruitment of inflammatory cells.
Liver inflammation might be terminated via the action of growth factors
like PDGF, TGF-ß1, and EGF derived from different cellular sources,
which are capable of diminishing I-CAM-1 and V-CAM-1 expression by HSC
(Figure 7)
. Although these regulatory loops must be studied in more
detail, the present report provides at least novel insights into the
mechanisms underlying hepatic inflammation and particularly into the
role of HSC as proinflammatory elements.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Department of Internal Medicine, Section of Gastroenterology and Endocrinology, University of Göttingen and by the Deutsche Forschungsgemeinschaft Sonderforschungsbereich 402, "Molekulare und Zelluläre Hepatogastroenterologie."
Accepted for publication October 1, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4ß1/VCAM-1 adhesion pathway in physiology and disease. Res Immunol 1993, 144:723-735[Medline]
: transcription in rat Kupffer cells and in vitro posttranslational processing based on a PCR-derived cDNA. Biol Chem 1992, 373:271-281
-skeletal and -cardiac actin but multicopy for ß- and -cytoskeletal genes: 3' untranslated regions are isotype specific but are conserved in evolution. Mol Cell Biol 1983, 3:1783-1791 on proliferation, activation and protein synthesis of rat hepatic stellate cells. J Hepatol 1997, 27:1067-1080[Medline]
, macrophage inflammatory protein-2, interleukin-1 ß, and interleukin-6 but not RANTES or transforming growth factor-ß 1 mRNA expression in acute lung inflammation. Am J Respir Cell Mol Biol 1994, 10:148-153[Abstract]
, other cytokines and adhesion molecules in human livers with alcoholic hepatitis. Virchows Arch A 1993, 423:169-176
messenger RNA. Exp Mol Pathol 1995, 62:42-51[Medline]
gene expression and the response to interferon in chronic hepatitis C. Hepatology 1996, 23:210-217[Medline]
and nitric oxide production in endotoxin-primed rats administered carbon tetrachloride. Life Sci 1995, 57:2273-2280[Medline]
and transforming growth factor-ß 1 in acute liver injury. Growth Factors 1989, 1:219-226[Medline]
-mRNA synthesis and distribution of tumor necrosis factor--mRNA synthesizing cells in rat liver during experimental endotoxemia. J Hepatol 1994, 20:122-128[Medline]
and interferon- in the liver biopsies of children with chronic liver disease. J Hepatol 1994, 21:816-821[Medline]
production by human hepatocytes in chronic viral hepatitis. J Exp Med 1994, 179:841-848This article has been cited by other articles:
 |
|
 |
|
  S. L. Friedman Hepatic Stellate Cells: Protean, Multifunctional, and Enigmatic Cells of the Liver Physiol Rev, January 1, 2008; 88(1): 125 - 172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Adrian, K. Poelstra, G. L. Scherphof, D. K. F. Meijer, A.-m. van Loenen-Weemaes, C. Reker-Smit, H. W. M. Morselt, P. Zwiers, and J. A. A. M. Kamps Effects of a New Bioactive Lipid-Based Drug Carrier on Cultured Hepatic Stellate Cells and Liver Fibrosis in Bile Duct-Ligated Rats J. Pharmacol. Exp. Ther., May 1, 2007; 321(2): 536 - 543. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Bergers and S. Song The role of pericytes in blood-vessel formation and maintenance Neuro-oncol, October 1, 2005; 7(4): 452 - 464. [Abstract] [PDF]
|
|
|
|
 |
|
 S. Hazra, S. Xiong, J. Wang, R. A. Rippe, V. Krishna, K. Chatterjee, and H. Tsukamoto Peroxisome Proliferator-activated Receptor {gamma} Induces a Phenotypic Switch from Activated to Quiescent Hepatic Stellate Cells J. Biol. Chem., March 19, 2004; 279(12): 11392 - 11401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Schwabe, R. Bataller, and D. A. Brenner Human hepatic stellate cells express CCR5 and RANTES to induce proliferation and migration Am J Physiol Gastrointest Liver Physiol, November 1, 2003; 285(5): G949 - G958. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Saile, N. Matthes, K. Neubauer, C. Eisenbach, H. El-Armouche, J. Dudas, and G. Ramadori Rat liver myofibroblasts and hepatic stellate cells differ in CD95-mediated apoptosis and response to TNF-alpha Am J Physiol Gastrointest Liver Physiol, August 1, 2002; 283(2): G435 - G444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Schwabe, B. Schnabl, Y. O. Kweon, and D. A. Brenner CD40 Activates NF-{{kappa}}B and c-Jun N-Terminal Kinase and Enhances Chemokine Secretion on Activated Human Hepatic Stellate Cells J. Immunol., June 1, 2001; 166(11): 6812 - 6819. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. J. Eng and S. L. Friedman Fibrogenesis I. New insights into hepatic stellate cell activation: the simple becomes complex Am J Physiol Gastrointest Liver Physiol, July 1, 2000; 279(1): G7 - G11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Friedman Molecular Regulation of Hepatic Fibrosis, an Integrated Cellular Response to Tissue Injury J. Biol. Chem., January 28, 2000; 275(4): 2247 - 2250. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Knittel, D. Kobold, J. Dudas, B. Saile, and G. Ramadori Role of the Ets-1 Transcription Factor during Activation of Rat Hepatic Stellate Cells in Culture Am. J. Pathol., December 1, 1999; 155(6): 1841 - 1848. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
