Expression of Cyclooxygenase 2 in Human Malignant

Melanoma
Carsten Denkert, Martin Köbel, Stefan Berger, et al.

Cancer Res 2001;61:303-308. Published online January 1, 2001.

Updated Version Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/61/1/303

Cited Articles This article cites 35 articles, 19 of which you can access for free at:
http://cancerres.aacrjournals.org/content/61/1/303.full.html#ref-list-1
Citing Articles This article has been cited by 17 HighWire-hosted articles. Access the articles at:
http://cancerres.aacrjournals.org/content/61/1/303.full.html#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at pubs@aacr.org.

Permissions To request permission to re-use all or part of this article, contact the AACR Publications
Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on April 26, 2011

Copyright © 2001 American Association for Cancer Research
[CANCER RESEARCH 61, 303–308, January 1, 2001]

Expression of Cyclooxygenase 2 in Human Malignant Melanoma

Carsten Denkert, Martin Köbel, Stefan Berger, Antje Siegert, Anja Leclere, Uwe Trefzer, and Steffen Hauptmann1
Institute of Pathology [C. D., M. K., S. B., A. S., A. L., S. H.] and Department of Dermatology and Allergy [U. T.], Charité Hospital, D-10117 Berlin, Germany

ABSTRACT Up to now, there is insufficient data supporting the function of

COX-2 in tumors of nonepithelial origin (14). Thus, it is not clear
Cyclooxygenase (COX)-2 is an inducible enzyme involved in production whether COX-2-induced tumor progression is restricted to epithelial
of prostaglandins in inflammatory processes. There is now increasing
cancer or if it represents a more general mechanism of tumorigenesis.
evidence that a constitutive expression of COX-2 plays a role in develop-
ment and progression of malignant epithelial tumors. In the present study
In the present study we investigated expression and function of
we investigated expression and function of COX-2 in malignant mela- COX-2 in malignant melanoma, a nonepithelial tumor characterized
noma. Expression of COX-2 was determined by immunohistochemistry in by a marked inflammatory stromal response.
28 cases of primary skin melanoma and 4 benign nevi. We show that
COX-2 was expressed in 26 cases (93%) of melanomas, with a moderate
to strong expression in 19 cases (68%). Benign nevi as well as normal MATERIALS AND METHODS
epithelium were negative in all cases. A constitutive expression of COX-2
Materials. The human melanoma cell lines A375 (15), MeWo (16), SK-
mRNA and protein was found in five melanoma cell lines (A375, MeWo,
Mel-13 (17), SK-Mel-28 (18), and IGR-37 (19) were cultured in DMEM
SK-Mel-13, SK-Mel-28, and IGR-37) by using Northern blot as well as
supplemented with 10% fetal bovine serum. NS-398 (Alexis, Grünberg, Ger-
immunoblotting. All melanoma cell lines produced prostaglandin (PG) E2
many) was dissolved in DMSO to a stock concentration of 100 mM. The cDNA
between 468 and 3500 pg/ml as determined by ELISA. Treatment with
fragment of human COX-2 as well as the mouse antihuman COX-2 mono-
NS-398 (50 ␮M), a specific inhibitor of COX-2, suppressed PGE2 produc-
clonal antibody used for immunoblotting was obtained from Cayman Bio-
tion of all melanoma cell lines by 50 –96%. The IC50 for inhibition of PGE2
chemicals (Ann Arbor, MI). The monoclonal antibody against human COX-2
production by NS-398 was determined as 4 ␮M, indicating that NS-398
(clone33) used for immunohistochemistry was from Transduction Laboratories
acts via inhibition of the COX-2 isoenzyme. We could show that prolifer-
(Lexington, KY).
ation of melanoma cell lines was not influenced by treatment with NS-398
Immunohistochemistry. Immunohistochemical examination was per-
in concentrations up to 100 ␮M. However, NS-398 reduced Matrigel
formed retrospectively on tissue samples taken for routine diagnostic purposes
invasion of all five malignant melanoma cell lines by 50 – 68%. Our results
from 32 patients who underwent excision of skin tumors between 1992 and
indicate that COX-2 is expressed in malignant melanomas and may be
1998 at the Department of Dermatology, Charité Hospital, Berlin. Twelve
involved in regulation of melanoma invasion. It remains to be investigated
cases were melanomas within the horizontal growth phase (in situ and Clark
whether selective inhibitors of COX-2 might be useful for prevention or
level 1 or 2), whereas the other 16 cases were melanomas within the vertical
treatment of malignant melanoma.
growth phase (Clark level 3, 4, or 5). Four cases of benign nevus cell nevi were
evaluated for comparison. Tissue samples were fixed in 4% neutral buffered
INTRODUCTION formaldehyde and embedded in paraffin. Routine H&E sections were per-
formed for histopathological evaluation. Immunohistochemical staining was
COXs2 are involved in control of inflammatory reactions and performed according to standard procedures. Briefly, slides were boiled in
catalyze the conversion of arachidonic acid to PGH2, which is the citrate buffer in a pressure cooker for 5 min and incubated with the monoclonal
precursor of prostanoids. Two COX isoenzymes have been described: COX-2 antibody (1:2000; Transduction Laboratories) overnight at 4°C, fol-
COX-1 is the constitutive form and is regarded as a housekeeping lowed by incubation with a biotinylated antimouse secondary antibody and the
multilink biotin-streptavidin-amplified detection system (Biogenex, San
gene, whereas COX-2 is highly inducible by inflammatory stimuli (1).
Ramon, CA). Staining was visualized using a Fastred chromogen system
The role of COX-2 has been extensively studied in colorectal cancer, (Immunotech, Hamburg, Germany). The intensity of the COX-2 immuno-
where this enzyme is expressed in adenomas and also is increased in staining in tumor cells as well as surrounding inflammatory cells was evaluated
carcinomas (2, 3). Epidemiological studies show that NSAIDs such as independently by two pathologists and scored semiquantitatively as ⫺, nega-
aspirin or sulindac reduce the incidence and mortality in colorectal tive; and ⫹, weak; ⫹⫹, moderate; and ⫹⫹⫹, strong positive.
carcinoma and several other types of cancer (4 –7). Furthermore, in Immunoblotting. Cells grown to confluency in 50-mm Petri dishes were
animal experiments inhibition of COX-2 reduced the incidence of lysed in 100 ␮l of 62.5 mM Tris-HCl (pH 6.8) containing 2% SDS, 10%
colon carcinoma in rats treated with chemical carcinogens (8) as well glycerol, 50 mM DTT, and 0.1% bromphenol blue. One hundred ␮g protein/
as in APC knockout mice (9). sample were loaded on a 10% polyacrylamide gel. Proteins were blotted onto
COX-2 is expressed in other carcinomas of the gastrointestinal tract nitrocellulose membranes (Biometra, Göttingen, Germany), washed in TBS,
and incubated in blocking buffer (1⫻ TBS, 0.1% Tween 20, 5% nonfat dry
as well, such as gastric or pancreatic adenocarcinomas (10). Addi-
milk) for 1 h at 21°C. Membranes were washed three times with TBS/0.1%
tionally, an enhanced expression of COX-2 has been observed in Tween 20 and incubated overnight at 4°C with a monoclonal anti-COX-2
well-differentiated hepatocellular carcinomas (11), well-differentiated antibody diluted 1:250 in TBS/0.1% Tween 20, followed by incubation with
adenocarcinomas of the lung (12), and in squamous carcinomas of the alkaline phosphatase-conjugated goat antirabbit secondary antibody (Tropix,
head and neck (13). Bedford, MA). Bands were visualized using the CDP star RTU luminescence
system (Tropix).
Received 2/2/00; accepted 11/1/00. Northern Hybridization. Total RNA was prepared with the RNeasy Kit
The costs of publication of this article were defrayed in part by the payment of page (Qiagen, Hilden, Germany). RNA samples (5 ␮g) were electrophoresed in 1%
charges. This article must therefore be hereby marked advertisement in accordance with agarose with 2.2 M formaldehyde and then blotted onto Hybond N⫹ mem-
18 U.S.C. Section 1734 solely to indicate this fact.
1 branes (Amersham, Braunschweig, Germany). After UV cross-linking (Hofer,
To whom requests for reprints should be addressed, at the Institute of Pathology,
Charité Hospital, D-10117 Berlin, Germany. San Francisco, CA), blots were hybridized in ExpressHyb hybridization solu-
2
The abbreviations used are: COX, cyclooxygenase; PG, prostaglandin; NSAID, tion (Clontech, Palo Alto, CA) with [␣-32P]dCTP-labeled, random-primed
nonsteroidal anti-inflammatory drug; TBS, Tris-buffered saline; GAPDH, glyceralde- cDNA probes using the megaprime DNA labeling system (Amersham). Blots
hyde-3-phosphate dehydrogenase; MTT, 3-(4,5-dimethylthiazole-2-yl)-2.5-diphenyltetra-
were exposed to Kodak Biomax films at ⫺70°C with intensifying screens. For
zolium bromide; XTT, 2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-2H-tetrazolium-5-
carboxanilide inner salt; PPAR, peroxisome proliferator-activated receptor; APC, standardization, hybridization with a cDNA probe for human GAPDH (Clon-
adenomatous polyposis coli. tech) was performed.
303

Downloaded from cancerres.aacrjournals.org on April 26, 2011

Copyright © 2001 American Association for Cancer Research
 COX-2 EXPRESSION IN MALIGNANT MELANOMA

PGE2 ELISA. 1 ⫻ 105 cells/well in 12-well plates were treated with or Table 1 Immunohistochemistry: COX-2 expression in benign nevi and malignant
without 50 ␮M NS 398 (Alexis) in DMEM plus 10% FCS. After 24 h, the melanoma
medium was replaced by 500 ␮l identical medium supplemented with 20 ␮M Clark Tumor thickness COX-2 COX-2
arachidonic acid (Sigma). The supernatants were harvested after 1 h and Case Histologya level (mm) tumorb stromab
centrifuged at 5000 rpm for 10 min before blocking the COX by addition of 10 1 BN ⫺ ⫺
␮g/ml indomethacin (Sigma). Samples were stored at ⫺80°C. 2 BN ⫺ ⫺
3 BN ⫺ ⫺
Concentration of PGE2 in cell culture supernatants was determined using a
4 BN ⫺ ⫺
specific ELISA (R&D Systems, Minneapolis, MN) according to the manufac- 5 MIS In situ ⫹⫹ ⫹⫹⫹
turer’s instructions. The concentration of PGE2 was estimated from the ab- 6 SSM 2 0.15 ⫹ ⫹⫹
sorbance of the calculated standard curve. The lower limit of sensitivity for 7 SSM 1 0.25 ⫹⫹ ⫹⫹⫹
8 SSM 2 0.28 ⫹⫹ ⫹⫹
detection of PGE2 was 36 pg/ml. The results are expressed as pg/ml per 105
9 LMM 2 0.33 ⫹⫹ ⫹⫹
cells. 10 SSM 2 0.34 ⫹ ⫹
Matrigel Invasion Assay. Transwells with polycarbonate membranes 11 SSM 2 0.34 ⫹⫹ ⫺
(8-␮m pores) in six-well tissue culture plates (Costar, Cambridge, MA) were 12 SSM 2 0.36 ⫹ ⫹⫹⫹
13 SSM 2 0.48 ⫹⫹ ⫹
coated with Matrigel (Becton Dickinson, Heidelberg, Germany) diluted in
14 SSM 3 0.49 ⫹⫹⫹ ⫹⫹⫹
DMEM (1 mg/ml; 675 ␮l/4.7 cm2) and incubated for 60 min at 37°C. 15 SSM 2 0.50 ⫹ ⫹⫹⫹
Afterward, membranes were washed once with DMEM. Cells were detached 16 SSM 2 0.54 ⫹⫹ ⫹
with trypsin/EDTA and seeded in the upper compartment of the transwell 17 LMM 2 0.60 ⫹⫹⫹ ⫹⫹
18 NM 4 2.05 ⫹⫹ ⫹⫹⫹
insert at a concentration of 2 ⫻ 105 cells/ml serum-free culture medium.
19 NM 4 2.2 ⫹⫹ ⫹
NS-398 (50 ␮M) or DMSO (0.05%) as control were added to the upper and the 20 NM 4 2.55 ⫹⫹ ⫹
lower compartments. In some experiments, PGE2 (0.05–100 nM) was added. 21 NM 4 2.64 ⫹ ⫹⫹⫹
After 72 h, cultures were incubated with 0.5 mg/ml MTT for 4 h, as described 22 NM 4 3.50 ⫺ ⫺
23 NM 4 3.69 ⫹⫹ ⫹⫹⫹
(20). Cells on the upper and lower surfaces of the transwell insert were
24 NM 4 3.72 ⫹ ⫹⫹⫹
removed separately, dissolved in DMSO, and measured using an ELISA 25 NM 4 4.00 ⫹ ⫹⫹
reader. Metabolization of MTT by the noninvasive cells on the upper surface 26 NM 4 4.50 ⫹⫹⫹ ⫹⫹
was found to be independent of treatment with NS-398 (data not shown) and 27 NM 5 5.00 ⫹⫹⫹ ⫹
28 NM 4 6.1 ⫹⫹⫹ ⫹⫹
was used to exclude an effect of the inhibitor treatment on cell proliferation
29 NM 4 6.75 ⫺ ⫹⫹
and cell viability. Invasion was expressed as percentage of reduction of 30 NM 5 7.20 ⫹⫹⫹ ⫹
invasion compared with control (DMSO-treated) cells. To exclude an effect of 31 NM 4 8.00 ⫹⫹ ⫺
NS-398 on MTT metabolization, NS-398 was added in some cases immedi- 32 NM 5 9.00 ⫹⫹ ⫺
a
ately before addition of MTT. This treatment did not change MTT metabo- BN, benign nevus; MIS, melanoma in situ; SSM, superficial spreading melanoma;
lization (data not shown). LMN; lentigo maligna melanoma; NM, nodular melanoma.
b
Staining: ⫹⫹⫹, high; ⫹⫹, moderate; ⫹, low; ⫺, negative expression.
Proliferation Assay. Cell proliferation was measured using an XTT test
(Boehringer-Mannheim, Mannheim, Germany). Cells were grown in 96-well
plates for 3 days in medium containing 10% fetal bovine serum and NS-398 in
concentrations between 0.1 and 100 ␮M. XTT metabolization was measured tumor thickness and COX-2 expression in tumor cells (Spearman’s
according to the manufacturer’s instructions. In additional experiments, cell rank correlation coefficient, 0.16) as well as stromal cells (correlation
proliferation was measured by direct cell counting using a CASY cell counter coefficient, ⫺0.26). Thus, no significant differences in COX-2 ex-
(Schärfe Systems, Reutlingen, Germany). These experiments gave results
pression in different stages of malignant melanoma were found.
similar to those of the XTT test (not shown).
Statistics. All data shown are from at least three independent experiments
Expression of COX-2 mRNA and Protein in Malignant Mela-
and are expressed as mean ⫾ SE. The statistical significance was determined noma Cell Lines. For additional investigation of expression of
using Student’s t test or Fisher’s exact test. P ⬍ 0.05 was considered as COX-2 in malignant melanoma, we determined the expression of
significant. For statistical evaluation, the SPSS software Version 8.0 was used. COX-2 mRNA and protein in five cell lines of malignant melanoma
(MeWo, SK-Mel-13, SK-Mel-28, IGR 37, A375). As shown in Fig. 2,
RESULTS constitutive expression of COX-2 mRNA with a transcript of 4.5 kb
was found in all cell lines. In Western blot analysis, all five melanoma
Expression of COX-2 Protein in Primary Tumors of Malignant cell lines expressed COX-2 protein with a size of Mr ⬃70,000 (Fig. 3).
Melanoma. We have investigated 28 cases of primary malignant Expression levels of COX-2 mRNA and protein were comparable
melanoma as well as 4 benign nevi for expression of COX-2 protein with the colon carcinoma cell line HT-29, which was used as a
by immunohistochemistry (Table 1). Twelve cases of primary mela- positive control (not shown).
noma were within the radial growth phase (in situ; Clark level 1 or 2), PGE2 Production of Malignant Melanoma Cells. As shown in
whereas 16 cases were within the vertical growth phase (Clark levels Fig. 4, all cell lines produced PGE2, with the highest production found
3–5). Expression of COX-2 in melanoma cells was found in 26 cases in A375 cells (3500 ⫾ 216 pg/ml) and the lowest in SK-Mel-13 cells
(93%), with moderate to strong intensity in 19 cases (68%). A gran- (468 ⫾ 18 pg/ml). In all cases, there was a strong reduction of PGE2
ular cytoplasmic staining pattern was found in melanoma cells as well production in cultures treated with NS-398 (Fig. 4). This reduction
as in inflammatory cells within the stroma, such as macrophages or was maximal in A375 cells (96% inhibition by NS-398) and minimal
lymphocytes (Fig. 1). Adjacent normal epithelium including the mela- in SK-Mel-28 cells (50% inhibition by NS-398). To estimate the IC50
nocytes, the underlying connective tissue as well as the benign nevi for inhibition of PGE2 production by NS-398, we incubated MeWo
were negative for COX-2 (Table 1, Fig. 1). cells with different concentrations of NS-398 (0.1–50 ␮M) and meas-
In nodular melanomas, there was a considerable heterogeneity of ured arachidonic acid-stimulated PGE2 production (Fig. 5). We found
COX-2 expression, with an enhancement of COX-2 expression in the a dose-dependent inhibition of PGE2 production by NS-398, with an
periphery of the tumor. The percentage of cases with a strong expres- estimated IC50 of 4 ␮M.
sion of COX-2 was higher in melanomas in the vertical growth (5 In additional experiments, we measured basal levels of PGE2 in
cases, 31%) than that in tumors in the radial growth phase (1 case, cells cultured for 24 h without further addition of arachidonic acid. In
8%). However, this relation was not statistically significant (Fisher’s these unstimulated cultures, PGE2 levels well above the detection
exact test; P ⫽ 0.19). We did not observe any correlation between limit of the ELISA were measured for all cell lines, ranging between
304

Downloaded from cancerres.aacrjournals.org on April 26, 2011

Copyright © 2001 American Association for Cancer Research
 COX-2 EXPRESSION IN MALIGNANT MELANOMA
Fig. 1. Immunohistochemistry: COX-2 expression in primary malignant melanoma.
Superficial spreading melanoma with immunoreactivity for COX-2 in melanoma cells as
well as stromal macrophages (A, arrowheads). Expression of COX-2 in nodular mela-
noma (B, arrowheads, stromal macrophages). C, benign nevus cell nevus showing no
immunoreactivity for COX-2.
321 ⫾ 28 pg/ml (IGR-37) and 138 ⫾ 23 pg/ml (SK-Mel-13; data not
shown). These basal levels of PGE2 were inhibited in all cell lines by
NS-398.
Inhibition of Matrigel Invasion by NS-398. To evaluate the in-
volvement of COX-2 in the invasion of malignant melanoma, a
305
Downloaded from cancerres.aacrjournals.org on April 26, 2011
Copyright © 2001 American Association for Cancer Research
Matrigel assay with or without addition of NS-398 was performed. In
all cell lines, invasion was reduced by NS-398 (50 ␮M; Fig. 6A).
Inhibition of Matrigel invasion by NS-398 in different cell lines was
between 50% (MeWo) and 66% (A375). To investigate the mecha-
nism of regulation of melanoma invasion by COX-2, we measured
Matrigel invasion after addition of exogenous PGE2. Exogenous
PGE2 (0.05–100 nM) neither enhanced Matrigel invasion nor reduced
Fig. 2. Expression of COX-2 mRNA in five human melanoma cell lines (MeWo, A375,
IGR 37, SK-Mel-13, SK-Mel-28) as detected by Northern blot. All of the melanoma cell
lines expressed COX-2 mRNA as a single transcript of 4.5 kb in levels comparable with
that of the colorectal carcinoma cell line HT-29 (not shown). Expression of GAPDH was
used for standardization of mRNA levels. One of three independent experiments is shown.
Fig. 3. Expression of COX-2 protein in five human melanoma cell lines (MeWo, A375,
IGR 37, SK-Mel-13, SK-Mel-28) as detected by immunoblotting. All of the melanoma cell
lines expressed COX-2 protein with a size of Mr ⬃70,000 in levels comparable with that
of the colorectal carcinoma cell line HT-29 (not shown). One of three independent
experiments is shown.
Fig. 4. Production of PGE2 by human melanoma cell lines. Cells were incubated with
(f) or without (䡺) NS-398 for 24 h. Medium was changed to medium containing 20 ␮M
arachidonic acid. Concentration of PGE2 in cell culture supernatants was determined by
specific ELISA after a 1-h incubation. All of the melanoma cell lines produced PGE2. In
all cases, there was a strong reduction of PGE2 produced in cultures treated with NS-398
(f). Mean and SE of three independent experiments are shown.
 COX-2 EXPRESSION IN MALIGNANT MELANOMA
Fig. 5. Modulation of PGE2 production of MeWo cells by different concentrations of
NS-398. MeWo cells were incubated with NS-398 for 24 h. Medium was changed to
identical medium containing 20 ␮M arachidonic acid. PGE2 production was measured by
specific ELISA in supernatants after 1 h. The IC50 for inhibition of PGE2 production by
NS-398 was estimated as 4 ␮M. Mean ⫾ SE of three independent experiments are shown.
Fig. 6. Matrigel invasion assay. Cells were cultured on transwell cell culture inserts
coated with Matrigel and treated with NS-398 (50 ␮M) or carrier alone (DMSO, 0.05%)
for 72 h. Treatment of melanoma cell lines with NS-398 reduced Matrigel invasion in all
of the cell lines (A). Treatment of MeWo cells with exogenous PGE2 (0.05–100 nM) did
not reduce NS-398-induced inhibition of invasion of MeWo cells (B). Invasion is ex-
pressed as percentage of reduction of invasion compared with control (DMSO-treated)
cells. Mean ⫾ SE of three independent experiments are shown. ⴱ, P ⬍ 0.05 (Student’s t
test).
NS-398-induced inhibition of invasion of MeWo cells (Fig. 6B).
Similar results were obtained for all other cell lines (data not shown).
Proliferation of Malignant Melanoma Is Not Changed by
NS-398. We measured proliferation of the five malignant melanoma
cell lines incubated with different concentrations of the specific
COX-2 inhibitor NS-398. As shown in Fig. 7 for MeWo cells, no
significant inhibition of cell proliferation was observed using concen-
trations of NS-398 between 0.1 and 100 ␮M. Similar results were
obtained in three independent experiments for each of the five mela-
noma cell lines (data not shown).
306
Downloaded from cancerres.aacrjournals.org on April 26, 2011
Copyright © 2001 American Association for Cancer Research
DISCUSSION
Several functions of inducible cyclooxygenase (COX-2) have been
described in the biology of various carcinomas: increased cell prolif-
eration (21), inhibition of apoptosis (22), stimulation of angiogenesis
(23), and inhibition of immunosurveillance (24). To our knowledge,
this is the first study investigating expression and function of COX-2
in human malignant melanoma. For mouse melanoma cells, it has
been shown that inhibitors of COXs suppressed invasion and MMP-2
production, but the expression and function of the COX-2 isoenzyme
was not investigated (25).
We show that COX-2 is expressed in the majority of primary
malignant melanomas, as well as in five human malignant melanoma
cell lines. These cell lines produced PGE2, which could be inhibited
by the specific COX-2 inhibitor NS-398. Inhibition of COX-2 did not
change proliferation of malignant melanoma but reduced Matrigel
invasion in all cell lines. Thus, in human malignant melanoma, there
seems to be an involvement of COX-2 in the regulation of tumor
invasion.
Studies on the function of COX-2 in other types of tumors support
this observation. For example, COX-2 expression in gastric carcinoma
was correlated with tumor invasion into lymphatic vessels as well as
metastasis into lymph nodes (26). Similar results have been shown for
pulmonary adenocarcinomas, where COX-2 expression was enhanced
in metastases compared with primary tumors (27). In colon carcinoma
cell lines, transfection with COX-2 resulted in increased Matrigel
invasion (28).
In our immunohistochemical studies, COX-2 expression was local-
ized to tumor cells as well as granulocytes and macrophages within
the stromal infiltrate. In nodular tumors, a higher expression of
COX-2 was usually found in the periphery of the tumor. This expres-
sion pattern might indicate that COX-2 expression is regulated by
interaction between stromal cells and tumor cells. Similar results have
been described for colon carcinoma as well as ulcerative colitis, where
an increased expression of COX-2 was found in subepithelial myofi-
broblasts (29) and macrophages (30).
To investigate the function of COX-2 in cell culture models, we
studied expression of COX-2 mRNA as well as of COX-2 protein in
five melanoma cell lines. All cell lines expressed levels of COX-2
comparable with the colon carcinoma cell line HT-29, which was used
as a positive control. The levels of PGE2 varied remarkably between
different cell lines. In all cases, PGE2 production could be inhibited by
the specific COX-2 inhibitor NS-398, indicating that COX-2 is the
major if not the only rate-limiting factor of PGE2 in melanoma cells.
Fig. 7. Proliferation of MeWo cells treated with NS-398. No significant inhibition of
cell proliferation was observed for MeWo cells by using concentrations of NS-398
between 0.1 and 100 ␮M. One of three independent experiments performed in triplicate
with MeWo cells is shown. Identical results were observed in three independent experi-
ments for each of the five melanoma cell lines (data not shown).
 COX-2 EXPRESSION IN MALIGNANT MELANOMA
For MeWo cells, we determined the IC50 for PGE2 inhibition by
NS-398 as approximately 4 ␮M, which is within the range of the
reported IC50 of NS-398 for inhibition of COX-2 (1.77 ␮M; Ref. 31),
whereas the IC50 for inhibition of COX-1 by NS-398 is 75 ␮M (31).
This suggests that the effect of NS-398 on melanoma PGE2 produc-
tion is mediated by inhibition of COX-2.
With regard to the function of COX-2 in melanoma biology, we did
not find a regulation of melanoma growth by the COX-2 inhibitor
NS-398. This is consistent with a recent publication, where it was
shown that COX inhibitors did not reduce the size of mouse mela-
noma tumors (32).
We were able to demonstrate that NS-398 could inhibit invasion of
malignant melanoma cells. The mechanisms involved in the inhibition
of invasion by NS-398 are not completely clear thus far. A PGE2-
mediated mechanism seems to be unlikely, because we have not been
able to modulate melanoma invasion with exogenous PGE2. Further-
more, NS-398-induced suppression of invasion could not be overcome
by addition of exogenous PGE2. This conclusion is additionally also
supported by the fact that the inhibitory effect on melanoma cell
invasion was seen at relatively high concentrations of NS-398 (50
␮M). We did not observe a significant inhibitory effect using lower
concentrations of NS-398 (data not shown). This suggests that the
inhibitory effect of NS-398 might be mediated by both COX-1 and
COX-2, although we found only low level expression of COX-1 in
melanoma cell lines by using immunoblotting (data not shown). Thus,
the molecular basis for control of invasion by NS-398 remains to be
clarified. The observation that treatment with PGE2 failed to reverse
the effects of NSAIDs has been made with colon cancer cells as well
(33, 34), raising the possibility for a non-COX target of NSAIDs. It
has been shown that the PPAR␦, which is regulated by APC, is
another target of the NSAIDs indomethacin and sulindac sulfide in
colon cancer for inhibition of tumorigenesis (35). However, inhibition
of PPAR␦ by COX-2-specific inhibitors such as NS-398 has not been
studied thus far. Because APC mutations have been described in
malignant melanoma cells (36), it should be interesting to investigate
the function of PPAR␦ in melanoma progression and invasion.
The development of new specific inhibitors of COX-2 leads to a
new concept for chemoprevention of cancer (37). Thus, it is important
to investigate the function of COX-2 in malignant tumors of nonepi-
thelial origin as well, to determine the indications for chemopreven-
tion. In malignant melanoma, where the primary tumor can be re-
moved at an early point in time but where metastases can arise despite
a relatively small tumor, patients may benefit from an antimetastatic
chemoprevention strategy. It will be important to determine on the
basis of the results of this study whether COX-2 expression is an
independent prognostic factor in malignant melanoma and if selective
inhibitors of COX-2 are useful in preventing or treating malignant
melanoma.
ACKNOWLEDGMENTS
We thank Ines Koch for her excellent technical assistance and Dr. Peter
Hufnagl for his help with the statistical analysis.
REFERENCES
1. O’Banion, M. K., Winn, V. D., and Young, D. A. cDNA cloning and functional
activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc. Natl.
Acad. Sci. USA, 89: 4888 – 4892, 1992.
2. Kargman, S. L., O’Neill, G. P., Vickers, P. J., Evans, J. F., Mancini, J. A., and Jothy,
S. Expression of prostaglandin G/H synthase-1 and -2 protein in human colon cancer.
Cancer Res., 55: 2556 –2559, 1995.
3. Sano, H., Kawahito, Y., Wilder, R., Hashiramoto, A., Mukai, S., Asai, K., Kimura, S.,
Kato, H., Kondo, M., and Hia, T. Expression of cyclooxygenase-1 and -2 in human
colorectal cancer. Cancer Res., 55: 3785–3789, 1995.
307
Downloaded from cancerres.aacrjournals.org on April 26, 2011
Copyright © 2001 American Association for Cancer Research
4. Thun, M. J., Namboodiri, M. M., and Heath, C. W., Jr. Aspirin use and reduced risk
of fatal colon cancer. N. Engl. J. Med., 325: 1593–1596, 1991.
5. Schreinemachers, D. M., and Everson, R. B. Aspirin use and lung, colon and breast
cancer incidence in a prospective study. Epidemiology, 5: 138 –146, 1994.
6. Taketo, M. M. Cyclooxygenase inhibitors in tumorigenesis (Part I). J. Natl. Cancer
Inst. (Bethesda), 90: 1529 –1536, 1998.
7. Taketo, M. M. Cyclooxygenase inhibitors in tumorigenesis (Part II). J. Natl. Cancer
Inst. (Bethesda), 90: 1609 –1615, 1998.
8. Kawamori, T., Rao, C. V., Seibert, K., and Reddy, B. S. Chemopreventive activity of
celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Can-
cer Res., 58: 409 – 412, 1998.
9. Oshima, M., Dinchuk, J. E., Kargman, S. L., Oshima, H., Hancock, B., Kwong, E.,
Trzaskos, J. M., Evans, J. F., and Taketo, M. M. Suppression of intestinal polyposis
in APC knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell, 87:
803– 809, 1996.
10. Tucker, O. N., Dannenberg, A. J., Yang, E. K., Zhang, F., Teng, L., Daly, J. M.,
Soslow, R. A., Masferrer, J. L., Woerner, B. M., Koki, A. T., and Fahey, T. J., III.
Cyclooxygenase-2 expression is upregulated in human pancreatic cancer. Cancer
Res., 59: 987–990, 1999.
11. Koga, H., Sakisaka, S., Ohishi, M., Kawaguchi, T., Taniguchi, E., Sasatomi, K.,
Harada, M., Kusaba, T., Tanaka, M., Kimura, R., Nakashima, Y., Nakashima, O.,
Kojiro, M., Kurohiji, T., and Sata, M. Expression of cyclooxygenase-2 in human
hepatocellular carcinoma: relevance to tumor dedifferentiation. Hepatology, 29: 688 –
696, 1999.
12. Wolff, H., Saukkonen, K., Anttila, S., Karjalainen, A., Vainio, H., and Ristimäki, A.
Expression of cyclooxygenase-2 in human lung carcinoma. Cancer Res., 58: 4997–
5001, 1998.
13. Chan, G., Boyle, J. O., Yang, E. K., Zhang, F., Sacks, P. G., Shah, J. P., Edelstein,
D., Soslow, R. A., Koki, A. T., Woerner, B. M., Masferrer, J. L., and Dannenberg,
A. J. Cyclooxygenase-2 expression is upregulated in squamous cell carcinoma of the
head and neck. Cancer Res., 59: 991–994, 1999.
14. Strausser, H. R., and Humes, J. L. Prostaglandin synthesis inhibition: effect on bone
changes and sarcoma tumor induction in balb/c mice. Int. J. Cancer, 15: 724 –730,
1975.
15. Raz, A., and Lotan, R. Lectin-like activities associated with human and murine
neoplastic cells. Cancer Res., 41: 3642–3647, 1981.
16. Carey, T. E., Takahashi, T., Resnick, L. A., Oettgen, H. F., and Old, L. J. Cell surface
antigens of human malignant melanoma: mixed hemabsorption assays for humoral
immunity to cultured autologous melanoma cells. Proc. Natl. Acad. Sci. USA, 73:
3278 –3282, 1976.
17. Livingston, P. O., Takeyama, H., Pollack, M. S., Houghton, A. N., Albino, A., Pinsky,
C. M., Oettgen, H. F., and Old, L. J. Serological responses of melanoma patients to
vaccines derived from allogeneic cultured melanoma cells. Int. J. Cancer, 31:567–
575, 1983.
18. Carey, T. E., Lloyd, K. O., Takahashi, T., Travassos, L. R., and Old, L. J. AU
cell-surface antigen of human malignant melanoma: solubilization and partial char-
acterization. Proc. Natl. Acad. Sci. USA, 76: 2898 –2902, 1979.
19. Aquaron, R., Dutoit, C., Reynier, M., and Aubert, C. Calmodulin content and
distribution in six human melanoma cell lines. Cell. Mol. Biol., 36: 85–92, 1990.
20. Imamura, H., Takao, S., and Aikou, T. A modified invasion-3-(4,5-dimethylthiazole-
2-yl)-2,5-diphenyltetrazolium bromide assay for quantitating tumor cell invasion.
Cancer Res., 54: 3620 –3624, 1994.
21. Sheng, H., Shao, J., Morrow, J. D., Beauchamp, R. D., and DuBois, R. N. Modulation
of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells.
Cancer Res., 58: 362–366, 1998.
22. Tsujii, M., and DuBois, R. N. Alterations in cellular adhesion and apoptosis in
epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell, 93:
705–716, 1995.
23. Tsujii, M., Kawano, S., Tsujii, S., Sawaoka, H., Hori, M., and DuBois, R. N.
Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell, 93:
705–716, 1998.
24. Huang, M., Stolina, M., Sharma, S., Mao, J. T., Zhu, L., Miller, P. W., Wollman, J.,
Herschman, H., and Dubinett, S. M. Non-small cell lung cancer cyclooxygenase-2-
dependent regulation of cytokine balance in lymphocytes and macrophages: up-
regulation of interleukin 10 and down-regulation of interleukin 12 production. Cancer
Res., 58: 1208 –1216, 1998.
25. Reich, R., and Martin, G. R. Identification of arachidonic acid pathways required for
the invasive and metastatic activity of malignant tumor cells. Prostaglandins, 51:
1–17, 1996.
26. Murata, H., Kawano, S., Tsujii, S., Tsujii, M., Sawaoka, H., Kimura, Y., Shiozaki, H.,
and Hori, M. Cyclooxygenase-2 overexpression enhances lymphatic invasion and
metastasis in human gastric carcinoma. Am. J. Gastroenterol., 94: 451– 455, 1999.
27. Hida, T., Yatabe, Y., Achiwa, H., Muramatsu, H., Kozaki, K., Nakamura, S., Ogawa,
M., Mitsudomi, T., Sugiura, T., and Takahashi, T. Increased expression of cyclooxy-
genase 2 occurs frequently in human lung cancers, specifically in adenocarcinomas.
Cancer Res., 58: 3761–3764, 1998.
28. Tsujii, M., Kawano, S., and Du Bois, R. N. Cyclooxygenase-2 expression in human
colon cancer cells increases metastatic potential. Proc. Natl. Acad. Sci. USA, 94:
336 –3340, 1997.
29. Shattuck-Brandt, R., Varilek, G. W., Radhika, A., Yang, F., Washington, M. K., and
DuBois, R. N. Cyclooxygenase 2 expression is increased in the stroma of colon
carcinomas from IL-10⫺/⫺ mice. Gastroenterology, 118: 337–345, 2000.
 COX-2 EXPRESSION IN MALIGNANT MELANOMA

30. Raab, Y., Sundberg, C., Hallgren, R., Knutson, L., and Gerdin, B. Mucosal synthesis
and release of prostaglandin E2 from activated eosinophils and macrophages in
ulcerative colitis. Am. J. Gastroenterol., 90: 614 – 620, 1995.
31. Barnett, J., Chow, J., Ives, D., Chiou, M., Mackenzie, R., Osen, E., Nguyen, B., Tsing,
S., Bach, C., Freire, J., Chan, H., Sigal, E., and Ramesha, G. Purification, character-
ization and selective inhibition of human prostaglandin G/H synthase 1 and 2
expressed in the baculovirus system. Biochim. Biophys. Acta, 1209: 130 –139, 1994.
32. Cahlin, C., Gelin, J., Delbro, D., Lönnroth, C., Doi, C., and Lundholm, K. Effect of
cyclooxygenase and nitric oxide synthase inhibitors on tumor growth in mouse tumor
models with or without cancer cachexia related to prostanoids. Cancer Res., 60:
1742–1749, 2000.
33. Hanif, R., Pittas, A., Feng, Y., Koutsos, M. I., Qiao, L., Stainano-Coico, L., Shiff,
S. I., and Rigas, B. Effects of nonsteroidal anti-inflammatory drugs on proliferation
and on induction of apoptosis in colon cancer cells by a prostaglandin-independent
pathway. Biochem. Pharmacol., 26: 237–245, 1996.
34. Piazza, G. A., Rahm, A. L., Krutzsch, M., Sperl, G., Paranka, N. S., Gross, P. H.,
Brendel, K., Burt, R. W., Alberts, D. S., Pamukcu, R., and Ahnen, D. J. Antineo-
plastic drugs sulindac sulfide and sulfone inhibit cell growth by inducing apoptosis.
Cancer Res., 14: 3110 –3116, 1995.
35. He, T-C., Chan, T. A., Vogelstein, B., and Kinzler, K. W. PPAR␦ is an APC regulated
target of nonsteroidal anti-inflammatory drugs. Cell, 99: 335–345, 1999.
36. Rubinfeld, B., Robbins, P., El-Gamil, M., Albert, I., Profiri, E., and Polakis, P.
Stabilization of ␤-catenin by genetic defects in melanoma cell lines. Science (Wash-
ington DC), 275: 1790 –1792, 1997.
37. Hong, W. K., and Sporn, M. B. Recent advances in chemoprevention of cancer.
Science (Washington DC), 278: 1073–1077, 1997.

308

Downloaded from cancerres.aacrjournals.org on April 26, 2011

Copyright © 2001 American Association for Cancer Research