| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Molecular Medicine Unit,*
the Department of
Histopathology,
and the Department of
Surgery,
University of Leeds, St. James’s
University Hospital, Leeds, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
COX-2 also plays an important role at an earlier stage of intestinal
tumorigenesis. Within intestinal adenomas of Min and
Apc
716 mouse models of familial
adenomatous polyposis, Cox-2 is localized to interstitial cells, which
have been identified as macrophages.17-19
Disruption of
Ptgs2 (the mouse COX-2 gene) and administration of the
selective Cox-2 inhibitor, MF-tricyclic, in the
Apc716 mouse have both been shown to
dramatically reduce intestinal adenoma development.17
Although sporadic colorectal adenoma development and progression are potential targets for colorectal cancer chemoprevention, the expression of COX-2 in human sporadic colorectal adenomas has received little attention. Studies of small numbers of sporadic and familial adenomatous polyposis adenomas have produced variable results with the expression of COX-2 demonstrated in between 0 to 90% of the adenomas.6,10,13,18,20 We, therefore, performed an immunohistochemical study of COX-2 protein expression and localization in a large series of human sporadic colorectal adenomas.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human Sporadic Colorectal Adenomas
Formalin-fixed, paraffin-embedded specimens of human sporadic colorectal adenomas and adjacent normal colorectal tissue were randomly selected from the histopathology archives at St. James’s University Hospital. All specimens had been obtained by endoscopic biopsy, polypectomy, or surgical resection during the preceding nine months. Adenomas from patients with inflammatory bowel disease or familial adenomatous polyposis were excluded from the study.
Adenoma tissue was also obtained immediately after endoscopic polypectomy, embedded in OCT compound (Merck Ltd., Poole, UK), and immediately snap-frozen in isopentane cooled in liquid nitrogen before storage at -70°C.
The age and gender of patients were noted. The size (maximum diameter in millimeters of polypectomy and surgical resection specimens), site (proximal or distal to the splenic flexure), histological type (tubular, tubulo-villous, or villous by WHO criteria21 ), and the highest histological grade of dysplasia (mild, moderate or severe) of adenomas were obtained from histopathology data files. Information including a history of NSAID (including low-dose aspirin use), oral corticosteroid use (defined as use at least once a month in 2 or more consecutive months during the previous 2 years), or family history of colorectal cancer (one or more first- or second-degree relatives diagnosed with colorectal cancer) was sought by a case note review (data ascertained in 95% of cases) and by a postal questionnaire (response rate 91%).
Immunohistochemistry
Formalin-fixed, paraffin-embedded sections (3 µm thick) were mounted on 3-aminopropyltriethoxysilane-coated glass slides. Sections were dewaxed in xylene (three times for 5 minutes), and rehydrated through a graded alcohol series (three times for 5 minutes). Endogenous peroxidase activity was blocked with 0.6% hydrogen peroxide in 100% methanol for 15 minutes at room temperature. After washing in water, nonspecific binding sites were blocked with 5% swine serum (DAKO Ltd., High Wycombe, UK) in phosphate-buffered saline (PBS) for 30 minutes at room temperature. We used affinity-purified rabbit polyclonal anti-human COX-2 IgG, (IBL, Gunma, Japan; 100 µg/ml),22,23 which was generated by immunization with a synthetic 17-mer peptide (TVKDTQAEMIYPPQVPE) corresponding to residues 251 to 267 of human COX-2. The primary antibody was diluted 1:25 in PBS and incubated with sections overnight at 4°C. After washing in PBS (two times for 10 minutes), sections were incubated with a 1:200 dilution of biotinylated swine anti-rabbit IgG (DAKO) for 25 minutes at room temperature. After further PBS washes, sections were incubated with streptavidin/biotin-horseradish peroxidase complex (DAKO) in 80 mmol/L Tris-HCl, pH 7.6, for 30 minutes at room temperature. Sections were visualized using 3,3'-diaminobenzidine tetrahydrochloride (0.7 mg/ml) and 0.05% hydrogen peroxide in 80 mmol/L Tris-HCl, pH 7.6, for 10 minutes. The sections were counterstained in Mayer’s hematoxylin for 3 minutes before dehydration, using a sequential alcohol and xylene series. The sections were mounted in diphenylxylene (BDH, Poole, UK).
Frozen sections (5 µm thick) were mounted on Superfrost Plus slides (BDH), fixed in 100% acetone for 2 minutes at room temperature, and air-dried for 10 minutes immediately before performing COX-2 immunohistochemistry as above.
The negative controls included omission of the primary antibody and antibody preadsorption with its cognate peptide (4 µg/ml, IBL) for 2 hours at 4°C. Human sporadic colorectal carcinomas (n = 3) were used as a positive tissue control.5 In addition, we confirmed the specificity of the primary antibody for COX-2 by Western blot analysis of purified ovine Cox-1 and Cox-2 (Cayman Chemical Co., Ann Arbor, MI) and whole-cell lysates of human umbilical vein endothelial cells in the absence (COX-2-negative) or presence of 20 ng/ml phorbol 12-myristate 13-acetate for 6 hours (COX-2-positive).
We also performed COX-2 immunohistochemistry on formalin-fixed, paraffin-embedded sections from the same adenoma series, using rabbit anti-mouse COX-2 antiserum (Cayman Chemical Co.), which we have described previously.19 The antiserum was generated by immunization with a synthetic 17-mer polypeptide (CY-SHSRLDDINPTVLIK), which corresponds to a C-terminal sequence in murine COX-2 (residues 584–598) with 80% homology to human COX-2. This antibody has previously been shown to recognize human COX-2 but not COX-1.24 Immunohistochemistry was performed as above except that the sections were placed in 10 mmol/L citrate buffer, pH 6.0, and heated to 80°C for 10 minutes in a microwave oven after the blocking of the endogenous peroxidase activity. The negative controls for this antibody, including antibody preadsorption, were performed as described.19
Adjacent sections were stained with mouse monoclonal anti-human CD68 IgG (clone KP1; DAKO), which recognizes mature tissue macrophages.25 The sections underwent antigen retrieval by pressure cooking (100°C, pressure 15 psi) in a 10 mmol/L citrate buffer, pH 6.0, for 60 seconds before cooling in tap water. The primary antibody, diluted 1:100 in PBS, was incubated with sections for 60 minutes at room temperature, and biotinylated rabbit anti-mouse IgG (DAKO) was used at a 1:200 dilution. The primary antibody was omitted as a negative control and human tonsil was used as a positive tissue control.26
Immunofluorescence
Colocalization of COX-2 and CD68 was performed using dual-labeling indirect immunofluorescence on frozen sections and using polyclonal anti-human COX-2 (IBL) and anti-CD68 antibodies. Nonspecific binding sites were blocked using 5% goat serum (Sigma Chemical Co., St Louis, MO) in PBS for 30 minutes at room temperature. COX-2 and CD68 antibodies (both diluted 1:25 in PBS) were simultaneously applied for 60 minutes at room temperature. After washes with PBS (four times for 5 minutes), sections were incubated with tetramethylrhodamine isothiocyanate-conjugated goat anti-rabbit immunoglobulin G (IgG; COX-2; Sigma) and fluorescein isothiocyanate-conjugated goat anti-mouse IgG (CD68; Sigma) for 60 minutes at room temperature at dilutions of 1:200 and 1:25, respectively. After washing in PBS four times for 5 minutes, sections were mounted in Vectashield (Vector Laboratories, Burlingame, CA) and visualized with a Zeiss Axioplan fluorescence microscope equipped with a dual filter for fluorescein isothiocyanate and tetramethylrhodamine isothiocyanate (Zeiss filter set 487924). The controls included the omission of one or both of the primary antibodies and the omission of one or both of the labeled secondary antibodies.
Semiquantitative Assessment of COX-2 and CD68 Immunostaining
COX-2 staining was assessed by two independent observers who were blinded to the origin of the sections. COX-2 protein expression in epithelial cells, interstitial cells immediately below the luminal surface of the adenoma (superficial interstitial cells), and interstitial cells deep within the body of the adenoma (deep interstitial cells) were scored separately on a scale of 0–3 based on both degree and intensity of staining (0, no staining; 1, small numbers, ie <10%, of COX-2-positive cells; 2, 10–50% of the cells were positive for COX-2; 3, >50% of cells were COX-2-positive with intense cellular staining). Data are expressed as the median COX-2 expression score and interquartile range (IQR). An interobserver concordance of COX-2 scores was 85% for superficial interstitial cells, 78% for deep interstitial cells, and 78% for epithelial cells.
The density of the CD68-positive macrophage infiltrate within adenomas was also measured. The mean number of CD68-positive cells counted in three random high-power fields (hpf) was scored on a scale of 1 to 3 (1, less than 25 CD68-positive cells/hpf; 2, 25–49 CD68-positive cells/hpf; 3, 50 or more CD68-positive cells/hpf).
Statistical Analysis
A logistic regression with forward conditional selection was performed to identify the factors that predicted COX-2 protein expression by adenomas. The following factors were included in the model: patient age, gender, NSAID/corticosteroid use, family history of colorectal carcinoma, adenoma site, adenoma size (for polypectomy and surgical specimens), histological type, grade of dysplasia and mode of excision (biopsy or polypectomy/surgical resection). The significance of the differences in COX-2 protein expression related to the adenoma site, histological type, grade of dysplasia, and patient age was tested using either a Mann-Whitney U test or Kruskal-Wallis one-way analysis of variance. The relationship between the interstitial cell COX-2 expression score and macrophage density was tested with the Spearman rank correlation coefficient. Statistical significance was assumed if the P value was less than or equal to 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Fifty-two formalin-fixed, paraffin-embedded adenoma specimens (14 endoscopic biopsies, 33 endoscopic polypectomies, 5 surgical polypectomies) were studied from 44 patients (24 male, 20 female, age range 44–84 years). Mean adenoma diameter was 11.9 ± 1.0 (SEM) mm. An examination of the sections stained with hematoxylin and eosin (H&E) revealed 27 (52%) tubular, 22 (42%) tubulo-villous, and 3 (6%) villous adenomas. Of the 52 adenomas, 20 (38%) exhibited mild dysplasia, 25 (48%) moderate dysplasia, and 7 (14%) severe dysplasia; 42 (81%) of the adenomas were located distal to the splenic flexure, and 10 (19%) were proximal to the splenic flexure. Of the 44 patients, 23 (46%) described a history of NSAID or corticosteroid use, with daily or alternate-day basis in 13 (30%). Five patients (11%) had a family history of colorectal cancer.
Six frozen, acetone-fixed specimens were studied from six patients (3 male, 3 female; age range 57–78 years). The mean adenoma size was 13.0 ± 2.9 (SEM) mm. An examination of the H&E-stained sections revealed four tubular adenomas and two tubulo-villous adenomas. Two adenomas exhibited mild dysplasia, two moderate dysplasia, and two severe dysplasia. Four adenomas were located distal to the splenic flexure, and two were proximal to the splenic flexure.
COX-2 Localization
Overall, COX-2 protein was detected in 40/52 (77%) human sporadic
colorectal adenomas. Immunoreactive COX-2 was localized to superficial
interstitial cells at the luminal surface of the adenoma in 39 (75%)
cases (Figure 1, A
-D). Aggregates of
COX-2-positive interstitial cells, which were located just below the
surface epithelium, were separated by interstitial tissue containing no
COX-2-positive cells (Figure 1A)
. The degree of superficial
interstitial cell COX-2 immunostaining varied between adenomas (Table 1)
. In 17 adenomas, widely scattered
small groups of COX-2-positive cells made up a small minority of the
superficial interstitial cell infiltrate of the adenoma. In other
adenomas (n = 10), COX-2-positive interstitial
cells accounted for the majority of the superficial interstitial cells
(scored 3; Figure 1, A and C
). COX-2-positive interstitial cells were
frequently observed at the tips of dysplastic villi (Figure 1, B
-D),
often beneath the flattened epithelium (Figure 1D)
. The close proximity
of COX-2-positive cells to capillaries was a prominent feature (Figure 1, B
-D).
|
|
. These
COX-2-positive cells were smaller and more widely separated than
superficial interstitial COX-2-positive cells. In the majority of
adenomas (n = 37), COX-2 was not detected in
epithelial cells. Dysplastic epithelial cells expressed COX-2 protein
in only 15 (29%) cases (Figure 1, C, D, and F)
. COX-2 staining was
noted in the superficial epithelial cells that were adjacent to
COX-2-positive interstitial cells (Figure 1, C, D, and F)
, but was also
observed in epithelial cells deep within adenomas. No immunoreactive
COX-2 protein was detected in paired samples (n
= 16) of histologically normal colorectal mucosa (Figure 1G)
.
The control sections confirmed the specificity of COX-2 staining
observed with polyclonal anti-human COX-2 IgG. The omission of the
primary antibody was associated with no positive staining.
Preadsorption of the antibody with its cognate peptide abolished
staining in interstitial cells (compare Figure 1
, panels H and J) and
in epithelial cells (data not shown). COX-2 immunohistochemistry
performed on acetone-fixed, frozen sections using both peroxidase- and
tetramethylrhodamine isothiocyanate-conjugated secondary antibodies
demonstrated identical COX-2 immunostaining to that obtained with
formalin-fixed sections (Figure 1, K and L)
. Immunohistochemistry with
murine COX-2 antiserum (n = 11) produced an
identical pattern of COX-2 staining to that seen with polyclonal
anti-human COX-2 IgG (Figure 1M)
. Staining with this antibody was also
abolished by antibody preadsorption with its cognate peptide.
Concordance between COX-2 expression scores for the two antibodies was
91% for superficial interstitial cells, 91% for deep interstitial
cells, and 82% for epithelial cells. A diffuse cytoplasmic staining of
cancer cells was consistently demonstrated in human colorectal
carcinoma specimens (Figure 1N)
as has been demonstrated by
others.5
Immunohistochemistry for CD68 (n = 26) revealed
large numbers of interstitial macrophages within adenomas. There was a
wide variation in the macrophage number in the adenomas examined (score
1, n = 11; 2, n = 8; 3,
n = 7). A proportion of the CD68-positive macrophage
population was observed at the tips of villi in a similar distribution
to COX-2-positive superficial interstitial cells. Furthermore,
COX-2-positive superficial interstitial cells had characteristic
features of macrophages (mononuclear cells with large ovoid nuclei;
Figure 1B
). Therefore, we proceeded to test whether COX-2 localized to
CD68-positive macrophages, using dual-labeling immunofluorescence. The
control sections showed no cross-reactivity between the primary and
secondary antibodies. Numerous interstitial CD68-positive cells (Figure 2A)
and a smaller number of interstitial
COX-2-positive cells (Figure 2B)
were observed by single
immunofluorescence. Observation with the dual filter demonstrated that
the COX-2 protein expression was restricted to a subgroup of
CD68-positive macrophages (Figure 2C)
. COX-2 localized to both
superficial and deep CD68-positive macrophages. Immunoreactive COX-2
was not detected in CD68-negative cells (Figure 2C)
.
|
A logistic regression analysis identified the adenoma site (P = 0.012) and histological type (P = 0.001) as significant independent predictors of superficial interstitial cell (macrophage) COX-2 protein expression. Age (P = 0.025) was the only significant independent predictor of COX-2 protein expression by deep interstitial cells. The degree of epithelial cell dysplasia did not predict epithelial cell COX-2 protein expression (P = 0.19, Kruskal-Wallis one-way analysis of variance).
The superficial interstitial cell COX-2 expression scores in adenomas distal to the splenic flexure were increased significantly (median COX-2 expression score 1; IQR 1–2.25; n = 42) compared with adenomas proximal to the splenic flexure (median score 0.5; IQR 0–1.25; n = 10; P = 0.03, Mann-Whitney U test). The superficial interstitial cell COX-2 expression scores for villous adenomas were significantly less (median score 0, n = 3) than corresponding values for both tubular (median score 1; IQR 1–2; n = 27) and tubulo-villous adenomas (median score 1; IQR 0–2; n = 22; P = 0.002, Mann-Whitney U test; combined tubular/tubulo-villous versus villous adenomas). COX-2 staining in deep interstitial cells in patients aged less than 65 years was increased significantly compared with those patients aged 65 years or more (P = 0.001, Mann-Whitney U test). The degree of macrophage infiltration in adenomas did not correlate with the COX-2 expression score for superficial and deep interstitial macrophages (r = 0.05; P = 0.79, Spearman rank correlation coefficient).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Existing immunohistochemical data on COX-2 expression in human and murine adenomas are conflicting.17-20,27 Therefore, we studied two different antibodies on both frozen and formalin-fixed sections and ensured that COX-2 staining with both antibodies was specific by preadsorption with the appropriate immunizing peptide. In addition, we demonstrated identical COX-2 immunoreactivity in colorectal cancer cells to that previously reported,5 and confirmed previous reports that histologically normal colonic mucosa does not express COX-2 protein.5-8,10,11 Moreover, the series of sporadic colorectal adenomas studied has wide clinical relevance as the adenomas had similar characteristics to those reported in the National Polyp Study.28 Because it remains unclear whether NSAIDs can attenuate expression, as well as inhibit the activity of COX-2,29-31 we also obtained data on NSAID use to ensure this did not confound our COX-2 expression data.
The finding that COX-2 is expressed predominantly by macrophages in
human sporadic colorectal adenomas is in agreement with data from
murine (Apc+/-) polyposis models.
We have previously reported macrophage-specific Cox-2 expression in
adenomas of Min mice.19
Similarly, Oshima et al used a
lacZ reporter under control of the Ptgs2 promoter
and localized Cox-2 expression to interstitial cells within adenomas of
Apc716 mice.17
By contrast,
Williams et al have described COX-2 immunostaining of epithelial cells
in Min mouse adenomas.27
However, this group did not use
controls confirming the specificity of staining for COX-2 in fixed
tissue sections that had been predigested with trypsin. The only
published immunohistochemical study on COX-2 localization in human
sporadic adenomas reported that COX-2 was localized to dysplastic
epithelial cells with only weakly positive staining of interstitial
cells.20
However, this study did not confirm antibody
specificity on formalin-fixed sections by antibody preadsorption and
demonstrated epithelial cell staining in normal colonic mucosa, which
conflicts with existing data.5-8,10,11
We did not
demonstrate COX-2 protein expression by other interstitial cell types,
such as the fibroblast, in our series, although others have reported
COX-2 expression by fibroblasts (identified by morphological criteria)
in adenomas of Apc716
mice32
and in Helicobacter pylori gastritis in
humans.33
A variable number of macrophages within adenomas expressed COX-2
protein, which could not be accounted for by differences in the density
of the macrophage infiltrate within individual adenomas. It is unclear
why only a variable proportion of the total macrophage population
within an adenoma expressed COX-2. However, a similar phenomenon has
been noted in a study of human colorectal cancers, in which tumor
necrosis factor 
expression was localized to less than 10% of
tumor-infiltrating macrophages.34
It is likely that the COX-2-expressing macrophages within adenomas
represent a population of activated macrophages. Human
macrophage/monocytes express COX-2 following activation by several
agents including bacterial lipopolysaccharide,35
interleukin-1ß,36
interferon-
, and tumor necrosis
factor-.37
The nature of an activating stimulus for
macrophages within adenomas remains open to speculation. A direct
activation of murine peritoneal macrophages by multiple tumor cell
lines has been reported,38
suggesting that activation may
be due to an antigenic stimulation from the adjacent adenomatous
epithelial cells. Alternatively, the passage of a luminal antigen such
as bacterial endotoxin across the dysplastic epithelial cell layer
could be responsible for the activation of superficial macrophages.
Coculture studies with human colorectal cancer cell lines and in
vivo studies in animals have demonstrated that macrophages can
stimulate tumor cell proliferation and migration.39-43
If
COX-2 does indeed play a role in the early stages of colorectal
carcinogenesis in man, COX-2 localization to interstitial macrophages
implies a paracrine-signaling pathway between macrophages and the
neighboring epithelial cells. Macrophages synthesize large quantities
of prostaglandin D2,44
derivatives
of which are capable of activating the transcription factor, peroxisome
proliferator-activated receptor ,45
thus providing an
auto- and/or paracrine mechanism whereby gene expression could be
altered. Other eicosanoids, such as PGE2, have
also been shown to directly increase the proliferation of human
colorectal cancer cell lines.46
In the only published
study of eicosanoid synthesis by macrophages isolated from adenomas, it
was reported that such macrophages do not synthesize excess
PGE2, compared with tissue macrophages from a
normal colon.47
However, this study only examined villous
adenomas, which, in our study, did not contain COX-2-positive
macrophages. Alternatively, COX-2-positive macrophages may play a role
in the adenoma progression via promotion of angiogenesis. COX-2
expression has been shown to stimulate angiogenesis in an in
vitro coculture model using Caco-2 cells.15
In this
regard, the close proximity demonstrated in our study of COX-2-positive
macrophages to capillaries is noteworthy.
COX-2-positive deep interstitial macrophages, noted in 17% of adenomas, were smaller than superficial COX-2-positive macrophages and probably represent a distinct macrophage population that can be isolated from the total tissue macrophage population of different tumors based on differential sedimentation characteristics.48 It is unclear why adenomas from older patients were significantly less likely to contain COX-2-positive deep interstitial macrophages. However, aged (greater than 65 years) human peripheral blood monocytes secrete decreased amounts of interleukin-1 after stimulation with lipopolysaccharide.49 This suggests that adenomas from older patients may contain fewer activated macrophages and, hence, a decreased macrophage COX-2 protein content.
COX-2 protein expression by macrophages was greater in the distal adenomas compared with the adenomas proximal to the splenic flexure. It is recognized that colorectal cancers exhibiting microsatellite instability have reduced COX-2 protein expression.50 It is possible that the predilection of mismatch repair-defective colon cancers, characterized by microsatellite instability, for the proximal colon could account for the differential COX-2 expression in adenomas demonstrated in our study.
We did not demonstrate macrophage COX-2 protein expression in villous adenomas, in contrast to tubular and tubulo-villous adenomas. Whether a causal relationship exists between the absence of macrophage COX-2 expression and the well-recognized increased malignant potential of villous adenomas remains to be determined.
Adenomatous epithelial cells expressed COX-2 protein in 29% of adenomas in our series. We found no relationship between the degree of dysplasia within adenomas and epithelial cell COX-2 protein expression, which is in agreement with data from Hao et al.20 In contrast, COX-2 is consistently expressed by the neoplastic epithelium in human colorectal carcinomas.5,11 This suggests that, in the majority of cases, up-regulation of COX-2 expression by epithelial cells is a relatively late event in colorectal carcinogenesis. Interestingly, COX-2 expression during development of esophageal adenocarcinoma from Barrett’s mucosa has similarities with our findings, in that COX-2 expression in the premalignant state (Barrett’s mucosa) is restricted to interstitial cells compared with established esophageal adenocarcinomas, which have been shown to contain COX-2-expressing malignant epithelial cells.51
A significant proportion of the adenomas (23%) that were studied did not express COX-2 protein detectable by our immunohistochemical technique. This has obvious implications for the development of selective COX-2 inhibitors as chemopreventative agents for colorectal cancer. The variable COX-2 expression by human sporadic colorectal adenomas should be taken into account when chemoprevention trials with selective COX-2 inhibitors are performed.
In summary, we have demonstrated that COX-2 protein is localized predominantly to superficial interstitial macrophages within human sporadic colorectal adenomas. If COX-2 plays a role at this stage of colorectal carcinogenesis, these data imply a paracrine signaling mechanism between macrophages and epithelial cells that warrants further investigation.
|
Footnotes |
|---|
Supported by a Royal College of Surgeons of England Clinical Research Fellowship (K. S. C.) and funding from Yorkshire Cancer Research (E. J. C. and G. H.) and the Medical Research Council of the United Kingdom (M. A. H.).
Accepted for publication October 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell 1996, 87:803-809[Medline]
gene expression in murine tissue macrophages. Mol Pharmacology 1997, 52:421-429 in human macrophages. J Immunol 1998, 161:2441–2448
release is induced by contact with some tumors. J Immunol 1990, 145:371-379[Abstract]
and promotes adipocyte differentiation. Cell 1995, 83:813-819[Medline]
This article has been cited by other articles:
 |
|
 |
|
  U. Dougherty, A. Sehdev, S. Cerda, R. Mustafi, N. Little, W. Yuan, S. Jagadeeswaran, A. Chumsangsri, J. Delgado, M. Tretiakova, et al. Epidermal Growth Factor Receptor Controls Flat Dysplastic Aberrant Crypt Foci Development and Colon Cancer Progression in the Rat Azoxymethane Model Clin. Cancer Res., April 15, 2008; 14(8): 2253 - 2262. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Samuelsson, R. Morgenstern, and P.-J. Jakobsson Membrane Prostaglandin E Synthase-1: A Novel Therapeutic Target Pharmacol. Rev., September 1, 2007; 59(3): 207 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Calviello, F. Resci, S. Serini, E. Piccioni, A. Toesca, A. Boninsegna, G. Monego, F. O. Ranelletti, and P. Palozza Docosahexaenoic acid induces proteasome-dependent degradation of {beta}-catenin, down-regulation of survivin and apoptosis in human colorectal cancer cells not expressing COX-2 Carcinogenesis, June 1, 2007; 28(6): 1202 - 1209. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.A. Hull, O.O. Faluyi, C.W.S. Ko, S. Holwell, D.J. Scott, R.J. Cuthbert, R. Poulsom, R. Goodlad, C. Bonifer, A.F. Markham, et al. Regulation of stromal cell cyclooxygenase-2 in the ApcMin/+ mouse model of intestinal tumorigenesis Carcinogenesis, March 1, 2006; 27(3): 382 - 391. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Shao, G. G. Sheng, R. C. Mifflin, D. W. Powell, and H. Sheng Roles of Myofibroblasts in Prostaglandin E2-Stimulated Intestinal Epithelial Proliferation and Angiogenesis Cancer Res., January 15, 2006; 66(2): 846 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Tanaka, A Tatsuguchi, S Futagami, K Gudis, K Wada, T Seo, K Mitsui, M Yonezawa, K Nagata, S Fujimori, et al. Monocyte chemoattractant protein 1 and macrophage cyclooxygenase 2 expression in colonic adenoma Gut, January 1, 2006; 55(1): 54 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Murtaugh, K.-n. Ma, B. J. Caan, C. Sweeney, R. Wolff, W. S. Samowitz, J. D. Potter, and M. L. Slattery Interactions of Peroxisome Proliferator-Activated Receptor {gamma} and Diet in Etiology of Colorectal Cancer Cancer Epidemiol. Biomarkers Prev., May 1, 2005; 14(5): 1224 - 1229. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Brown and R. N. DuBois COX-2: A Molecular Target for Colorectal Cancer Prevention J. Clin. Oncol., April 20, 2005; 23(12): 2840 - 2855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Ulrich, J. Whitton, J.-H. Yu, J. Sibert, R. Sparks, J. D. Potter, and J. Bigler PTGS2 (COX-2) -765G > C Promoter Variant Reduces Risk of Colorectal Adenoma among Nonusers of Nonsteroidal Anti-inflammatory Drugs Cancer Epidemiol. Biomarkers Prev., March 1, 2005; 14(3): 616 - 619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. T. Soumaoro, H. Uetake, T. Higuchi, Y. Takagi, M. Enomoto, and K. Sugihara Cyclooxygenase-2 Expression: A Significant Prognostic Indicator for Patients With Colorectal Cancer Clin. Cancer Res., December 15, 2004; 10(24): 8465 - 8471. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Adegboyega, O. Ololade, J. Saada, R. Mifflin, J. F. Di Mari, and D. W. Powell Subepithelial Myofibroblasts Express Cyclooxygenase-2 in Colorectal Tubular Adenomas Clin. Cancer Res., September 1, 2004; 10(17): 5870 - 5879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Hull, S. C.W. Ko, and G. Hawcroft Prostaglandin EP receptors: Targets for treatment and prevention of colorectal cancer? Mol. Cancer Ther., August 1, 2004; 3(8): 1031 - 1039. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Huycke and H. R. Gaskins Commensal Bacteria, Redox Stress, and Colorectal Cancer: Mechanisms and Models Experimental Biology and Medicine, July 1, 2004; 229(7): 586 - 597. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Quidville, N. Segond, E. Pidoux, R. Cohen, A. Jullienne, and S. Lausson Tumor Growth Inhibition by Indomethacin in a Mouse Model of Human Medullary Thyroid Cancer: Implication of Cyclooxygenases and 15-Hydroxyprostaglandin Dehydrogenase Endocrinology, May 1, 2004; 145(5): 2561 - 2571. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Nath, K. Kashfi, J. Chen, and B. Rigas Nitric oxide-donating aspirin inhibits {beta}-catenin/T cell factor (TCF) signaling in SW480 colon cancer cells by disrupting the nuclear {beta}-catenin-TCF association PNAS, October 28, 2003; 100(22): 12584 - 12589. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Gallo, N. Schiavone, L. Papucci, I. Sardi, L. Magnelli, A. Franchi, E. Masini, and S. Capaccioli Down-Regulation of Nitric Oxide Synthase-2 and Cyclooxygenase-2 Pathways by p53 in Squamous Cell Carcinoma Am. J. Pathol., August 1, 2003; 163(2): 723 - 732. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. W. J. de Leng, A. M. Westerman, M. A. J. Weterman, F. W. M. de Rooij, H. v. Dekken, A. F. P. M. de Goeij, S. B. Gruber, J. H. P. Wilson, G. J. A. Offerhaus, F. M. Giardiello, et al. Cyclooxygenase 2 Expression and Molecular Alterations in Peutz-Jeghers Hamartomas and Carcinomas Clin. Cancer Res., August 1, 2003; 9(8): 3065 - 3072. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Einspahr, R. S. Krouse, J. M. Yochim, P. V. Danenberg, K. D. Danenberg, A. K. Bhattacharyya, M. E. Martinez, and D. S. Alberts Association between Cyclooxygenase Expression and Colorectal Adenoma Characteristics Cancer Res., July 15, 2003; 63(14): 3891 - 3893. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. DuBois Evaluation of the Whole Prostaglandin Biosynthetic Pathway in Lung Cancer Clin. Cancer Res., May 1, 2003; 9(5): 1577 - 1578. [Full Text] [PDF]
|
|
|
|
|
|
Y. Zhu, P. Hua, and P. Lance Cyclooxygenase-2 Expression and Prostanoid Biogenesis Reflect Clinical Phenotype in Human Colorectal Fibroblast Strains Cancer Res., January 15, 2003; 63(2): 522 - 526. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
