| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Department of Veterans Affairs MedicalCenter,
Long Beach, California; the Universityof California,
Irvine, California; andthe Department of Surgery I,*
Oita Medical University, Oita, Japan
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
NSAIDs inhibit cyclooxygenase (COX), a rate-limiting enzyme in prostaglandin synthesis.11 Two isoforms of this enzyme are known: COX-1, constitutively expressed in most tissues, is responsible for the physiological production of prostaglandins; and COX-2, induced by cytokines, mitogens, and endotoxins in inflammatory cells, is responsible for the increased production of prostaglandins during inflammation.12 Early studies examining the effect of NSAIDs on experimental gastric ulcer healing1-3 have been conducted with the use of nonselective NSAIDs (eg, indomethacin) that inhibit both COX-1 and COX-2 enzymes.13 However, more recent studies indicate that experimental gastric ulceration induces COX-2, but not COX-1, expression14,15 and that COX-2 selective NSAIDs delay gastric ulcer healing similarly to nonselective NSAIDs.16 The expression of COX in ulcerated esophageal mucosa has not been studied and it is not known whether the COX-2 selective inhibitors affect esophageal ulcer healing.
Various growth factors, including epidermal growth factor and hepatocyte growth factor (HGF), have been implicated in the stimulation of epithelial proliferation during gastric ulcer healing.9,17,18 Suppression of HGF production has been suggested as a key factor involved in the inhibitory action of NSAIDs on gastric ulcer healing.19,20 However, the role of endogenous HGF in the healing of esophageal ulcers remains unexplored. In regard to the esophagus, previous studies demonstrated that exogenous HGF is the most potent stimulator of proliferation and restitution of esophageal epithelial cells in vitro, suggesting that it might be involved in the repair process of esophageal mucosal damage.21,22
In previous studies we have demonstrated that the extracellular signal-regulated kinases (ERKs) and their upstream kinase, Raf-1, which mediate the mitogenic effects of growth factors, are activated during gastric ulcer healing6-8 and that interruption of this signaling pathway dramatically delays the healing process.6 However, the signaling pathways involved in esophageal ulcer healing, and stimulation of esophageal epithelial cell proliferation essential for the healing, remain unknown.
The present study was aimed to: 1) explore molecular events associated with esophageal ulcer healing: expression of COX-1, COX-2, HGF and its receptor c-Met, and ERK2 phosphorylation levels and activity; and, 2) determine whether the selective COX-2 inhibitor, celecoxib, affects esophageal ulcer healing, ulceration-triggered cell proliferation and the above stated molecular events.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
, anti-c-Met p140, anti-ERK2
antibodies and monoclonal mouse anti-pERK antibody were purchased from
Santa Cruz Biotechnology (Santa Cruz, CA), monoclonal mouse
anti-proliferating cell nuclear antigen (PCNA) antibody and
3,3'-diaminobenzidine tetrahydrochloride were purchased from DAKO
(Carpinteria, CA). [
-32P]ATP was purchased
from Dupont NEN Research Products (Boston, MA). Keratinocyte basal
media was purchased from Clonetics (San Diego, CA). All other chemicals
were purchased from Sigma Chemical Company (St. Louis, MO). Induction of Esophageal Ulcers
This study was approved by the Subcommittee for Animal Studies of
the Long Beach Department of Veterans Affairs Medical Center. Male
Sprague-Dawley rats (weighing 225 to 250 g) were used in the
experiments. Rats were kept individually in wire-bottom cages with free
access to a standard rat chow (Rodent Diet No.8504; Harlan Teklad,
Madison, WI) and water. The animal room was illuminated on 12-hour
light-dark cycles. Room temperature was kept at 18 to 22°C and
humidity at 60 to 70%. Rats fasted for 12 hours underwent laparotomy
under ketamine-xylazine anesthesia (40 mg/kg body wt of ketamine and 5
mg/kg body wt of xylazine, intraperitoneally). Esophageal ulcers were
induced by modification of the method described by Tsuji and
colleagues.23
In brief, the stomach was pulled down and
the intra-abdominal esophagus was exposed. The length of esophagus,
brought below the diaphragm by this procedure, was 
1 cm providing
sufficient space for application of acetic acid. After careful
separation of vessels and nerves, 100% acetic acid (30 µl) was
applied through a polyethylene tube (3-mm inner diameter) to the
anterior wall of intra-abdominal esophagus for 3 minutes. The area was
then washed with isotonic saline and the abdomen was closed. The rats
undergoing the operation as described above but without application of
acetic acid (sham operation) served as controls. All rats survived the
scheduled study design.
Study Design
To evaluate the effects of ulceration on the studied
parameters, rats with esophageal ulcers and sham-operated (SO) rats
were euthanized 3 or 7 days after ulcer induction. To evaluate the
effects of celecoxib on ulcer healing, rats with esophageal ulcers were
treated once daily with either 10 mg/kg of celecoxib
(4-[5-(4-methylphenyl)-3-(trifluoro-methyl)-1H-pyrazol-1-yl]benzenesulfonamide;
Pfizer Inc., New York, NY) or its vehicle (1 ml of 1% carboxymethyl
cellulose sodium salt) starting 3 days after ulcer induction. Celecoxib
was given at 10 mg/kg because this dose has been shown to
significantly inhibit the COX-2-induced generation of prostaglandin
E2 in the rat air pouch without affecting
whole-blood thromboxane synthesis, an index of COX-1
activity.24
Treatments were given intragastrically for 2
or 4 days, and rats were euthanized 5 and 7 days after ulcer induction.
The starting of treatment (day 3) and ending (day 7) time points were
chosen because our previous preliminary studies demonstrated that
esophageal ulcers are fully developed by day 3 after application of
acetic acid and that 60% of ulcers are completely re-epithelialized
by day 9.10
A one-cm long segment of the lower esophagus
(including ulcer if present) was excised, opened longitudinally, and
photographed. The ulcer area was measured using a computerized
video-image analysis system (Image 1/FL; Universal Imaging Corp.,
Westchester, PA). Esophageal tissue samples including ulcer and
immediately adjacent tissue (ulcerated tissue) or esophageal tissue
from SO rats (normal tissue) were snap-frozen in liquid nitrogen, and
stored at -80°C for RNA isolation and protein extraction or fixed in
10% formalin. In addition, a 1-cm long segment of the esophagus distal
to the resected segment was excised and fixed in formalin for
immunohistochemical staining.
Determination of COX-1, COX-2, and c-Met mRNAs by RT-PCR
RNA was isolated using the RNeasy Mini Kit according to the manufacturer’s instructions. Reverse transcription (RT) and polymerase chain reaction (PCR) were performed to determine mRNA levels. RT was performed using a GeneAmp RNA PCR kit and a DNA thermal cycler (Perkin Elmer, Norwalk, CT), which were also used for PCR. Total RNA (0.3 µg) was used as the template to synthesize complementary DNA (cDNA) with 2.5 U of Moloney murine leukemia virus reverse transcriptase in 10 µl of buffer containing 10 mmol of Tris-HCl, pH 8.3, 50 mmol KCl, 5 mmol random hexamer, 1.4 U of ribonuclease inhibitor. RT was performed at 42°C for 15 minutes. The resulting cDNA was used as a template for subsequent PCR. The specific primer set for rat COX-1 was 5'-ACGCCCTCATTCACCCATTT-3' (sense) and 5'-CACGGACGCCTGTTCTACGG-3' (anti-sense) and for COX-2 was 5'-TGGTGCCGGGTCTGATGATG-3' (sense) and 5'-GCAATGCGGTTCTGATACTG-3' (anti-sense), and the sizes of amplified fragments were 561 bp for COX-1 and 253 bp for COX-2, respectively.25 The primers for rat c-Met were 5'-GCGAACTAATTCACTGCCCA-3' (sense) and 5'-GCATCTGTGTTGTGTACGGT-3' (antisense), and the size of the amplified fragment was 242 bp.26 The PCR for ß-actin was used as a positive control and an internal standard. The specific primer set for rat ß-actin was purchased from Clontech Laboratories, Inc., Palo Alto, CA. The PCR was performed in 50 µl of buffer containing 10 mmol Tris-HCl, pH 8.3, 2 mmol MgCl2, 50 mmol KCl, 0.2 mmol each of deoxyribonucleoside triphosphates, 0.4 µmol of each primer, 2 U of Taq DNA polymerase. For the amplification of COX-1, COX-2, and rat c-Met cDNAs, 35 cycles of 1 minute at 94°C for denaturing, 1 minute at 55°C for annealing, and 2 minutes at 72°C for extension were performed. Nine-µl aliquots of the PCR products were subjected to electrophoresis on a 1.25% agarose gel, and the DNA was visualized by ethidium bromide staining. Location of the products and their sizes were determined using a 100-bp ladder (Life Technologies, Inc., Gaithersburg, MD). The gel was then photographed under UV transillumination. For the quantitative assessment of the PCR products, a computerized video analysis system (Image-1/FL, Universal Imaging Corp.) was used. The results are expressed as target cDNA/ß-actin ratio.
Protein Extraction
Esophageal tissues were homogenized with a Polytron homogenizer (Kinematica, Littau, Switzerland) in a lysis buffer containing 62.5 mmol ethylenediaminetetraacetic acid, 50 mmol Tris, pH 8.0, 0.4% deoxycholic acid, 1% Nonidet P-40, 0.5 mg/ml leupeptin, 0.5 mg/ml pepstatin, 0.5 mg/ml aprotinin, 0.2 mmol phenylmethylsulfonyl fluoride, and 0.05 mmol aminoethyl benzene sulfonyl fluoride. The homogenates were then centrifuged at 14,000 rpm for 10 minutes at 4°C. The protein concentration of the supernatant was determined by the bicinchoninic acid protein assay kit.
Determination of COX-1, COX-2, HGF, and c-Met Protein Levels by Western Blotting
Equal amounts of protein (0.15 mg) were subjected to sodium
dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to
nitrocellulose membranes. The membranes were incubated with specific
antibodies at room temperature for 1 hour. The membranes were washed
and incubated with corresponding IgG peroxidase conjugates at room
temperature for 1 hour. The signal of the bound antibody was visualized
using enhanced chemiluminescence Western blotting detection reagents.
Protein expression was measured using a computerized video analysis
system (Image-1/FL, Universal Imaging Corp.). COX-1 and COX-2 protein
levels were determined using monoclonal mouse anti-COX-1 antibody and
polyclonal rabbit anti-COX-2 antibody (Cayman Chemical Co.) diluted
1:1000. For the determination of HGF and c-Met protein expression,
polyclonal rabbit anti-HGF- antibody (Santa Cruz Biotechnology,
Santa Cruz, CA) and polyclonal rabbit anti-c-Met p140 antibody (Santa
Cruz Biotechnology) were used at 1:250 dilution.
Determination of ERK2 Activity and Phosphorylation Levels
ERK2 activity was determined as described
previously.6
Briefly, 50 µg of protein from tissue
lysates was added to a conjugate of protein A Sepharose and 1 µg of
polyclonal rabbit anti-ERK2 antibody and mixed at 4°C for 2 hours.
The conjugates were then pelleted by centrifugation and washed four
times. After the final wash, buffer was removed completely and 40 µl
of protein kinase assay mixture (10 mmol/L HEPES, pH 7.5, 10 mmol/L
MgCl2, 50 µmol/L ATP, 30 µg myelin basic
protein, and 4 µCi [-32P]ATP) were added
to each sample. The samples were incubated at 30°C for 20 minutes and
the reaction was terminated by the addition of sodium dodecyl
sulfate-polyacrylamide gel electrophoresis sample buffer. The samples
were then electrophoresed on 15% acrylamide gels. After
electrophoresis, the gels were stained with Coomassie brilliant blue
and dried. The gels were autoradiographed; the myelin basic protein
bands were cut out and the radioactivity was counted in a scintillation
counter. The activity was expressed as pmol of
[-32P]ATP incorporated into 1 mg of MBP
protein (pmol/mg). ERK2 and phosphorylated ERK (pERK) protein levels
were determined by Western blot analysis using polyclonal rabbit
anti-ERK2 antibody and monoclonal mouse anti-pERK antibody. The ERK2
phosphorylation levels were expressed as a percentage of total ERK2
protein levels.
Localization of COX-2 and c-Met Protein Expression by Immunohistochemical Staining
Immunohistochemical staining with specific antibodies was performed to determine localization of COX-2 and c-Met expression in normal and ulcerated esophageal sections. Deparaffinized sections were incubated overnight at 4°C with polyclonal rabbit anti-COX-2 antibody diluted 1:300 in phosphate-buffered saline (PBS) or polyclonal rabbit anti-c-Met antibody diluted 1:50 in PBS. After washing, sections were then incubated for 30 minutes with rabbit fluorescein-conjugated IgG diluted 1:100 in PBS. Immunofluorescence signal was evaluated under a Nikon Optiphot epifluorescence microscope with B-filter composition (Nikon, Garden City, NY).
Evaluation of Esophageal Epithelial Cell Proliferation
Expression of PCNA in formalin-fixed paraffin-embedded esophageal tissue sections was determined by the enhanced polymer one-step staining method.27 Deparaffinized sections were incubated with monoclonal mouse anti-PCNA antibody for 1 hour at room temperature. The color was developed with 3,3'-diaminobenzidine tetrahydrochloride (DAKO) and the sections were counterstained with Mayer’s hematoxylin. Coded specimens were evaluated quantitatively under x400 microscopic magnification by two investigators unaware of the code. Cell nuclei that stained dark brown were considered as labeled. Labeled cells were counted in the epithelium above the edge of interrupted muscularis mucosa at each ulcer margin corresponding to 700 µm of mucosal section length, in the epithelium distant from the ulcer, and in normal esophageal epithelium of SO rats. The length of the basement membrane was measured on the photographed images of corresponding hematoxylin and eosin (H&E) sections using a computerized video analysis system (Image-1/FL, Universal Imaging Corp.) and the number of PCNA-labeled cells per 100-µm length of the basement membrane was calculated. The results are expressed as percentage of increase in the number of labeled cells in the epithelium of the ulcer margin over the number of labeled cells in the epithelium distant from the ulcer. Sections from six rats per group were evaluated and the mean ± SD was calculated.
Assessment of Angiogenesis in Granulation Tissue at the Ulcer Bed
To assess angiogenesis, enhanced polymer one-step immunostaining with monoclonal mouse anti-factor VIII-related antigen antibody (DAKO) that visualizes endothelial cells of vessels was used. Five days after ulcer induction, only a few endothelial cells forming microvessels were present in granulation tissue at the ulcer bed. Therefore, microvessels were counted only in sections obtained 7 days after ulcer induction. Coded specimens were evaluated quantitatively under x200 microscopic magnification by two investigators unaware of the code. Microvessels with distinct lumen were counted in granulation tissue below the regenerating epithelium of the ulcer margin at each side. The results are expressed as a number of microvessels (mean ± SD) per x200 microscopic field. Sections from six rats per group were evaluated.
Effects of Exogenous HGF on Esophageal Mucosal ERK2 Phosphorylation Levels
Rats (n = 18) were euthanized and a 1-cm long segment of the lower esophagus was excised and opened longitudinally. The esophageal mucosa was stripped from its muscle layers by sharp dissection as described previously.28 Explants consisted of squamous epithelium, lamina propria, and muscularis mucosa as determined by H&E staining. Esophageal mucosal explants were incubated in serum-free keratinocyte basal medium at 37°C with 5% CO2 and 95% air in a humidified incubator in the presence of 10 µmol/L or 100 µmol/L of SC-236 (Searle, Skokie, IL), a structural analog of celecoxib,29 or its vehicle (dimethyl sulfoxide) for 6 hours. SC-236 was used because of its better solubility compared to celecoxib. Then explants were treated with either human recombinant HGF (100 ng/ml) or its vehicle (PBS) and were incubated for an additional 30 minutes. Protein extraction and Western blotting for pERK2 and ERK2 were performed as described above.
Statistical Analysis
Values are expressed as the mean ± SD. Student’s t-test was used to determine the statistical significance of the differences. One-way analysis of variance followed by Bonferroni correction was used for multiple comparisons. A P value of <0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Three days after ulcer induction, COX-2 mRNA and protein levels in
ulcerated esophageal tissue were increased 
2.5-fold and
threefold, respectively, versus normal esophageal tissue
(Figures 1 and 2)
. COX-2 protein levels in ulcerated
esophageal tissue were decreased at 7 days versus 3 days
after ulcer induction, but remained elevated versus normal
tissue. COX-1 mRNA and protein levels were not significantly affected
by esophageal ulceration (Figures 1 and 2)
.
|
|
In normal esophageal epithelium, faint COX-2 signal was detected
predominantly in the cells of the basal zone (Figure 3A)
. Three days after ulcer induction, a
strong COX-2 fluorescence signal was present in all epithelial cells
constituting the ulcer margin (Figure 3B)
. In the normal esophageal
epithelium of sham-operated rats, c-Met signal was predominantly
localized to the membranes of cells in basal and stratified zones
(Figure 4A)
. Three days after ulcer
induction, c-Met signal was present in all epithelial cells
constituting ulcer margin; cells in the stratified zone displayed
strong membrane and cytoplasmic staining, whereas, basal cells showed
mainly diffuse cytoplasmic staining (Figure 4B)
.
|
|
The mean ± SD area of the ulcers 3 days after ulcer
induction was 7.1 ± 1.5 mm2. Five days
after ulcer induction, the mean ulcer area was slightly but
significantly decreased (versus 3 days) in
vehicle-treated but not in celecoxib-treated rats (Figure 5)
. Seven days after ulcer induction, the
mean ulcer area was significantly decreased (versus 3
days) in both vehicle-treated and celecoxib-treated rats. However, the
mean ulcer area in celecoxib-treated rats was nearly twofold larger
than in vehicle-treated rats, demonstrating that celecoxib treatment
significantly delays esophageal ulcer healing.
|
Seven days after ulcer induction, in the esophageal epithelium
distant from the ulcer, PCNA staining was present only in cells of the
basal zone (Figure 6, A and B)
.
Epithelium of the ulcer margin in rats treated with vehicle showed a
dramatic increase in the number of cells expressing PCNA (Figure 6C)
.
This increase was significantly diminished in rats treated with
celecoxib (Figure 6D)
. The quantitative data are shown in Figure 7
. The
number of PCNA-labeled epithelial cells in normal esophageal mucosa
(mucosa of SO rats) was not significantly different between the lower
(corresponding to the ulcer site in rats with ulcers) and upper
(corresponding to the area distant from the ulcer) esophageal segments
(18.5 ± 1.1 versus 17.9 ± 2.2, respectively;
P = 0.501). Three days after ulcer induction, the
number of PCNA-labeled epithelial cells in the mucosa distant from the
ulcer (17.2 ± 2.1) was not significantly
(P = 0.55) different from that in the mucosa of
the upper esophageal segment of SO rats. Three days after ulcer
induction, the number of PCNA-labeled epithelial cells at the ulcer
margin was increased to 180% over that in the epithelium distant from
the ulcer. In vehicle-treated rats, this increase reached 390% at 7
days after ulcer induction (Figure 7)
. In
rats treated with celecoxib, the number of PCNA-labeled epithelial
cells in the epithelium of the ulcer margin was also increased
(versus epithelium distant from the ulcer), however,
this increase was significantly reduced (versus
vehicle-treated rats) both 5 and 7 days after ulcer induction (Figure 7)
.
|
|
Seven days after ulcer induction, the number of microvessels in granulation tissue at the ulcer bed was slightly, but not significantly, lower in celecoxib-treated rats versus vehicle-treated rats (21.7 ± 1.7 versus 23.7 ± 3.0, respectively; P = 0.18).
Esophageal Ulceration Induces c-Met mRNA and Protein Expression: Effects of Celecoxib
Three days after ulcer induction, c-Met mRNA levels in ulcerated
tissue were increased twofold versus normal esophageal
tissue (Figure 8)
. In vehicle-treated
rats, c-Met mRNA levels in ulcerated tissue remained high at 5 days and
were decreased at 7 days versus 3 days after ulcer induction
(Figure 8)
. Celecoxib significantly reduced (versus
vehicle) c-Met mRNA expression in ulcerated tissue both 5 and 7 days
after ulcer induction (Figure 8)
. Western blot analysis demonstrated
that HGF protein levels in ulcerated tissue were significantly
increased versus normal esophageal tissue 3 days after ulcer
induction (Figure 9)
. In vehicle-treated
rats, HGF protein levels remained high 5 and 7 days after ulcer
induction (Figure 9)
. Celecoxib (versus vehicle) did
not significantly affect the ulceration-induced increase in HGF protein
levels at either time point. In vehicle-treated rats, c-Met protein
levels in ulcerated tissue were significantly increased at both 5 and 7
days versus 3 days. Celecoxib treatment
(versus vehicle treatment) significantly reduced
c-Met protein levels in ulcerated tissue both 5 and 7 days after ulcer
induction.
|
|
Three days after ulcer induction, ERK2 phosphorylation levels and
ERK2 activity were increased more than fourfold and fivefold,
respectively, in ulcerated tissue versus normal esophageal
tissue (Figure 10)
. In vehicle-treated
rats, ERK2 phosphorylation levels and ERK2 activity remained high at
both 5 and 7 days after ulcer induction. In celecoxib-treated rats,
ERK2 phosphorylation levels and ERK2 activity in ulcerated tissue were
significantly reduced versus vehicle-treated rats at both 5
and 7 days. ERK2 total protein levels were not significantly different
between any of the groups studied, demonstrating that celecoxib does
not reduce ERK2 protein expression but inhibits its activation
directly.
|
Dose response studies using 1 to 100 ng/ml of HGF revealed that
the 100-ng/ml dose significantly increased ERK2 phosphorylation levels
in esophageal mucosal explants. A thirty-minute incubation with 100
ng/ml of HGF significantly increased ERK2 phosphorylation levels in
esophageal mucosal explants (Figure 11)
. For ex vivo studies we
used a structural analog of celecoxib, SC-236, because of its better
solubility compared to celecoxib. Pretreatment with both 10 µmol/L
and 100 µmol/L SC-236 completely blocked the HGF-induced increase in
ERK2 phosphorylation levels. Basal ERK2 phosphorylation levels were not
significantly altered by 10 µmol/L of SC-236, however, they were
slightly, but significantly reduced by 100 µmol/L of SC-236. ERK2
total protein levels were not significantly different between the
studied groups.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Epithelial cell proliferation is a crucial component of gastric ulcer healing because it provides the cells necessary to fill the mucosal defect and to restore the mucosa within the scar.5 It has been suggested that inhibition of COX activity and the resulting inhibition of prostaglandin synthesis may play a role in NSAID-induced inhibition of epithelial cell proliferation during gastric ulcer healing.4 Prostaglandins have been demonstrated to induce HGF expression in gastric fibroblasts in vitro19 and, conversely, inhibition of prostaglandin synthesis by indomethacin significantly reduced HGF production in these cells.20 Therefore, it has been suggested that effects of NSAIDs on gastric ulcer healing may be mediated by suppression of prostaglandin-dependent HGF expression.20 However, whether NSAIDs affect HGF protein levels in vivo in the stomach or in the esophagus remains unknown. In the present study, HGF protein levels were increased during esophageal ulcer healing but were not affected by treatment with celecoxib, suggesting that effects of celecoxib on esophageal epithelial proliferation are not mediated via inhibition of HGF expression. On the other hand, there is evidence indicating an interaction between HGF and COX-2. Previously, we have demonstrated that HGF induces COX-2 expression in gastric epithelial cells in vitro.25 A recent in vivo study has also demonstrated that exogenous HGF up-regulates COX-2 expression in ulcerated gastric mucosa in rats.30 Moreover, the acceleration of gastric ulcer healing by HGF was significantly attenuated by selective COX-2 inhibitors, suggesting that COX-2 may mediate effects of HGF on gastric ulcer healing.30 However, the molecular mechanisms involved in this phenomenon remain unclear.
HGF elicits its biological effects by binding to and activating its receptor, c-Met,31 which is present on the majority of epithelial cells, including esophageal epithelial cells.22 The present study demonstrated that c-Met is expressed in esophageal epithelial basal and squamous cells and that its expression is induced by esophageal ulceration. Furthermore, we show that celecoxib significantly suppresses c-Met mRNA and protein expression during esophageal ulcer healing. Previous in vitro studies demonstrated that c-Met mRNA is inducible and that its expression is increased in response to HGF in cell lines derived from human lung adenocarcinoma and glioblastoma.32,33 Because inhibition of ERK activity completely prevented HGF-mediated c-Met induction in human glioblastoma cells,33 it is possible that celecoxib-induced inhibition of ERK2 activity is involved in the down-regulation of c-Met mRNA expression during esophageal ulcer healing. Because c-Met mediates the mitogenic effects of HGF, suppression of its expression may be one of the mechanisms by which celecoxib inhibits epithelial proliferation during esophageal ulcer healing.
HGF binding and activation of its receptor, c-Met, triggers activation of ERKs and cell proliferation in many cell types.34-36 Our previous studies demonstrated that HGF activates ERK2 in gastric epithelial cells.25 Inhibition of enhanced epithelial proliferation during gastric ulcer healing by NSAIDs has been well documented in experimental models and in humans.3,4,16 Previous studies demonstrated that gastric ulceration activates epidermal growth factor/mitogen-activated protein kinase (ERK2) mitogenic signaling pathway,6 but no studies have evaluated the effects of NSAIDs on this pathway during gastric or esophageal ulcer healing. Our present study showed that ERK2 is also activated during esophageal ulcer healing. Moreover, it demonstrated for the first time that celecoxib suppresses ERK2 activity up-regulated by esophageal ulceration and that a structural analog of celecoxib, SC-236, inhibits basal and HGF-induced ERK2 phosphorylation in esophageal mucosal explants. Taken together, these results indicate that COX-2 inhibitors may directly or indirectly (via interference with upstream events of ERK signaling pathway) inhibit ERK2 activation in esophageal mucosa. Because activation of the ERK2 pathway is essential for cell proliferation, the inhibitory action of celecoxib on esophageal epithelial proliferation is likely mediated by suppression of the ulceration-triggered ERK2 activation.
Studies of experimental gastric ulcers demonstrated that NSAIDs,
including COX-2-selective inhibitors, in addition to inhibition of
epithelial cell proliferation at the ulcer margin, also interfere with
angiogenesis (formation of new capillary blood vessels) in granulation
tissue at the ulcer bed.3,16
In the present study,
celecoxib, slightly (8%) but not significantly, inhibited
angiogenesis in granulation tissue during esophageal ulcer healing.
This finding suggests that inhibition of epithelial proliferation may
be more important for celecoxib-induced delay in esophageal ulcer
healing than inhibition of angiogenesis. This finding is in contrast to
NSAID-induced inhibition of gastric ulcer healing, whereas NSAIDs
markedly inhibit angiogenesis and suggests that angiogenesis may be
more important for healing of gastric ulcers than for healing of
esophageal ulcers. However, more detailed sequential studies
specifically assessing angiogenesis, as well as using angiogenic growth
factors are necessary to determine the precise role of angiogenesis in
esophageal ulcer healing.
In conclusion, esophageal ulceration triggers increases in esophageal epithelial proliferation; COX-2, HGF, and c-Met expressions; and ERK2 activity. A selective COX-2 inhibitor, celecoxib, delays healing of experimental esophageal ulcers and suppresses ulceration-induced increase in epithelial cell proliferation, c-Met induction, and ERK2 activation.
|
Footnotes |
|---|
Supported by the Medical Research Service of Veterans Administration, Minority Initiative, Merit Review, and Research Enhancement Award Program (to A. S. T.).
Dr. Hirofumi Kawanaka was a visiting scientist from Department of Surgery II, Kyushu University, Fukuoka, Japan; Dr. Woo S. Moon was a visiting scientist from Department of Pathology, Chonbuk National University, Chonju, South Korea; and Dr. Imre L. Szabo was a visiting scientist from Department of Internal Medicine, University of Pecs, Pecs, Hungary.
Accepted for publication November 30, 2001.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Chai, M. Norng, A. S Tarnawski, and J. Chow A critical role of serum response factor in myofibroblast differentiation during experimental oesophageal ulcer healing in rats Gut, May 1, 2007; 56(5): 621 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T Hayakawa, Y Fujiwara, M Hamaguchi, T Sugawa, M Okuyama, E Sasaki, T Watanabe, K Tominaga, N Oshitani, K Higuchi, et al. Roles of cyclooxygenase 2 and microsomal prostaglandin E synthase 1 in rat acid reflux oesophagitis Gut, April 1, 2006; 55(4): 450 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Miura, A. Tatsuguchi, K. Wada, H. Takeyama, Y. Shinji, T. Hiratsuka, S. Futagami, K. Miyake, K. Gudis, Y. Mizokami, et al. Cyclooxygenase-2-regulated vascular endothelial growth factor release in gastric fibroblasts Am J Physiol Gastrointest Liver Physiol, August 1, 2004; 287(2): G444 - G451. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. A. Bondesen, S. T. Mills, K. M. Kegley, and G. K. Pavlath The COX-2 pathway is essential during early stages of skeletal muscle regeneration Am J Physiol Cell Physiol, August 1, 2004; 287(2): C475 - C483. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. WARNER and J. A. MITCHELL Cyclooxygenases: new forms, new inhibitors, and lessons from the clinic FASEB J, May 1, 2004; 18(7): 790 - 804. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Wei, L. Wang, Y. He, H. Q. Xiong, J. L. Abbruzzese, and K. Xie Celecoxib Inhibits Vascular Endothelial Growth Factor Expression in and Reduces Angiogenesis and Metastasis of Human Pancreatic Cancer via Suppression of Sp1 Transcription Factor Activity Cancer Res., March 15, 2004; 64(6): 2030 - 2038. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Tanigawa, T. Watanabe, M. Hamaguchi, E. Sasaki, K. Tominaga, Y. Fujiwara, N. Oshitani, T. Matsumoto, K. Higuchi, and T. Arakawa Anti-inflammatory effect of two isoforms of COX in H. pylori-induced gastritis in mice: possible involvement of PGE2 Am J Physiol Gastrointest Liver Physiol, January 1, 2004; 286(1): G148 - G156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Caputo, C. Tuccillo, B. A. Manzo, R. Zarrilli, G. Tortora, C. D. V. Blanco, V. Ricci, F. Ciardiello, and M. Romano Helicobacter pylori VacA Toxin Up-Regulates Vascular Endothelial Growth Factor Expression in MKN 28 Gastric Cells through an Epidermal Growth Factor Receptor-, Cyclooxygenase-2-dependent Mechanism Clin. Cancer Res., June 1, 2003; 9(6): 2015 - 2021. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. A. Callejas, A. Fernandez-Martinez, A. Castrillo, L. Bosca, and P. Martin-Sanz Selective Inhibitors of Cyclooxygenase-2 Delay the Activation of Nuclear Factor kappa B and Attenuate the Expression of Inflammatory Genes in Murine Macrophages Treated with Lipopolysaccharide Mol. Pharmacol., March 1, 2003; 63(3): 671 - 677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Baatar, M. K. Jones, K. Tsugawa, R. Pai, W. S. Moon, G. Y. Koh, I. Kim, S. Kitano, and A. S. Tarnawski Esophageal Ulceration Triggers Expression of Hypoxia-Inducible Factor-1{alpha} and Activates Vascular Endothelial Growth Factor Gene : Implications for Angiogenesis and Ulcer Healing Am. J. Pathol., October 1, 2002; 161(4): 1449 - 1457. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
