Although recent studies have demonstrated that cyclooxygenase (COX)-2 is overexpressed in various cancers including gastric cancer, the mechanisms underlying the contribution of COX-2 to tumorigenesis and tumor promotion still remain unclear. To determine the role of COX-2, we investigated the COX-2 expression, the prostaglandin (PG) levels, and the microvessel density in 42 patients with primary gastric adenocarcinoma. COX-2 protein was overexpressed in 31 (74%) of 42 gastric cancers based on an immunoblot analysis. The intensity of COX-2 expression was found to significantly correlate with lymph node involvement. The COX-2 overexpressed cases showed significantly elevated levels of prostaglandin E2 (PGE2) in cancer tissues in comparison with the normal gastric mucosa by an immunoassay(201 ± 90 versus 161 ± 57 ng/mg protein; P < 0.05). However, the COX-2 overexpression was not related to the levels of thromboxane B2 and 6-keto-prostaglandin F. The density of microvessel immunostained with CD34 was significantly higher in patients demonstrating COX-2 overexpression than in those without such expression (63 ± 21 versus 45 ± 17/200 ×; P < 0.01). Our data thus suggested COX-2 overexpression to be associated with increased PGE2biosynthesis and angiogenesis in gastric cancer, which indicates that COX-2 may play a role in the development of gastric cancer.

COX2 is a key enzyme in PG biosynthesis (1). Two isoforms of COX have been identified: constitutively expressed COX-1 and mitogen-inducible COX-2 (2, 3, 4). Recently, an enhanced expression of COX-2, but not COX-1, has been found in cases of colon cancer (5, 6, 7, 8). Various epidemiological studies have previously revealed that the use of NSAIDs can reduce the risk of colon cancer (9, 10, 11). Because NSAIDs are known to inhibit COX, their beneficial effect in colon cancer is considered to be associated with COX-2 overexpression in this disease. COX-2 overexpression has been recently observed in many tumors other than the colon, such as the lung, the breast, and the esophagus (12, 13, 14, 15, 16). In gastric cancer, we and others have demonstrated an enhanced expression of COX-2 protein and mRNA in most cancer tissues compared with the accompanying normal mucosa (15, 16).

Recent studies have demonstrated that COX-2 could affect carcinogenesis via several different mechanisms. COX-2-mediated PG biosynthesis has been suggested to be involved in the development of cancer based on elevated levels of PGs, especially PGE2,in cancer tissues (17, 18, 19). COX-2 has been also reported to induce angiogenesis, which might be essential for tumor growth (20). COX-2 may be related to the development of gastric cancer as well, however, its association with PG biosynthesis and angiogenesis still remains unclear. To determine the role of COX-2 expression in gastric cancers, we examined the PG levels and microvessel density in patients with gastric cancer, and then compared the findings with the expression of COX-2 protein.

Patients.

Forty-two patients undergoing surgery for primary gastric cancer at the National Defense Medical College Hospital from 1993 to 1998 were examined. Of these, 29 were male and 13 were female. The mean age was 62 years (range, 31–86). Paired samples of cancer tissue and normal gastric mucosa were obtained from each patient at the time of surgery. The samples were immediately frozen in liquid nitrogen and were stored at −80°C. The remaining tissue specimens were fixed in 10% formalin and embedded in paraffin for a routine histological examination and immunohistochemical analysis.

Western Blot Analysis for COX-2.

Immunoblot analysis was performed as described previously (15). Briefly, cell lysates (20 μg/lane) were separated on 10% SDS polyacrylamide gel and then were electrophoretically transferred to a polyvinylidene difluoride membrane. COX-2 protein was detected by a rabbit polyclonal IgG (Immunobiological Laboratories,Fujioka, Japan) and visualized by the enhanced chemiluminescence system(Amersham, Arlington Heights, IL). The density of the bands was quantitated using the NIH image software package (Version 1.61). The intensity of COX-2 expression was judged by the ratio of its expression in cancer tissue (C) to its expression in paired normal gastric mucosa (N), and a ratio (COX-2 C:N) of more than 1.0 was considered to indicate an overexpression of COX-2 (15).

Immunoassay for PGs.

The levels (ng/mg protein) of PGE2,TXB2, and 6-k-PGF were examined using an immunossasy kit (Amersham) in paired protein samples of gastric cancer and normal mucosa.

Immunohistochemistry.

To detect microvessels on paraffin embedded samples,anti-CD34 (a monoclonal antibody to a transmembrane protein found on immature endothelial cells) was used (DAKO, Kyoto, Japan). Four-μm-thick sections were deparaffinized, microwaved for 15 min for antigen retrieval, immersed in 0.3% hydrogen peroxide for 30 min, and then immersed in normal goat serum for 30 min. The slides were incubated with anti-CD34 overnight at 4°C, and then stained by the standard streptavidin-biotin immunoperoxidase method. The areas containing a large number of microvessels or “hot spots” were identified at low magnification (×40) using a light microscope. Once hot spots were recognized, microvessels per field were counted at×200. From the 10 fields counted, the highest number of microvessels was used to determine the density (21, 22).

Statistical Analysis.

Differences between the groups were analyzed by theχ 2 test or Welch’s t test. Pearson’s correlation coefficient (r) was tested by the F test. A P < 0.05 was considered to be statistically significant.

COX-2 Protein Expression.

The expression of COX-2 protein was up-regulated (COX-2 C:N >1.0) in 31 (74%) of 42 gastric carcinomas by an immunoblot analysis(Fig. 1). The ratio of COX-2 C:N was statistically higher in the patients with lymph node involvement than in those without (P <0.05). No significant differences in the ratio were found in either the histological type or the tumor invasion (Table 1).

PG Levels.

The levels of PGE2 were higher in the cancer tissues than in the normal gastric mucosa (P < 0.05). The differences in the levels of TXB2 and 6-k-PGF between the normal and the cancer tissue specimens were not statistically significant. No significant association was seen between the levels of any of the three PGs and the histological type, depth of invasion, or lymph node involvement (Table 2).

The cases with COX-2 overexpression showed significantly higher levels of PGE2 in the cancer tissues than in the normal gastric mucosa (P < 0.05; Table 2), whereas the cases without COX-2 overexpression did not. The levels of TXB2 and 6-k-PGF did not differ between the normal and cancer tissue specimens in either the patients with COX-2 overexpression or those without (Table 2).

Microvessel Density.

The mean microvessel density of all of the cases was 58 (SD, 20), with a range of 31–121 (Fig. 2). No significant relationship was observed between the microvessel density and the histological type, depth of invasion, or lymph node involvement(data not shown).

The microvessel density was significantly higher in patients with COX-2 overexpression than in those without (63 ± 21 versus45 ± 17; ×200 microfield; P < 0.01). A significant correlation was observed between the ratio of COX-2 C:N and the microvessel density (P < 0.0001; Fig. 3,A). However, the PGE2 levels in the cancer tissue specimens showed no correlation with the microvessel density (Fig. 3 B).

In the present study, we provided a profile of an enhanced expression of COX-2 protein and elevated levels of PGE2 in human gastric cancer tissues compared with the normal mucosa. Increased levels of PGE2in the cancer tissue specimens were prominent in patients with COX-2 overexpression but not in those without. We previously demonstrated by immunohistochemical staining (15) that the main source of increased COX-2 protein in the cancer tissue was the cancer cells themselves. These findings suggest that gastric cancer cells overexpressing COX-2 may promote PGE2biosynthesis. PGE2 shows a potent immunosuppression effect by inhibiting the T-cell or natural killer cell activity (23, 24, 25, 26). Elevated PGE2 levels may, thus, provide a selective advantage for cancer cell survival, which may lead to the development of gastric cancer.

TXA2 has been reported to facilitate tumor metastasis (27, 28), whereas PGI2 has been shown to be an antimetastatic agent (29). The levels of TXA2 (assayed as its product TXB2) and PGI2(assayed as its product 6-k-PGF) were not associated with COX-2 expression in this study. The effects of TXA2 and PGI2, thus, may not be involved in the role that COX-2 plays in the progression of gastric cancer.

The present study demonstrated that the intensity of COX-2 expression correlated with the metastatic involvement of the lymph nodes. In contrast, the levels of PGE2 did not correlate with lymph node involvement. COX-2-expressing colon cancer cells have been reported to enhance the metastatic potential by activating metalloproteinase (30). COX-2 overexpression may thus enhance the lymphatic invasion in gastric cancer by a mechanism different from PGE2 biosynthesis, possibly because of the activation of metalloproteinase.

Angiogenesis is well recognized to be essential for the growth of solid tumors (31, 32). In the present study, we confirmed that the microvessel density correlated with the intensity of COX-2 expression. There was no significant correlation between the PGE2 levels and the microvessel density in the cancer tissue specimens, which, thus, suggests that PGE2 biosynthesis—mediated via the COX pathway—may not act directly on endothelial cells in gastric cancer. COX-2 has been shown in colon cancer cell lines to stimulate angiogenesis by inducing such angiogenic factors as vascular endothelial growth factor and transforming growth factor β (20). These mechanisms may, therefore, play a role in the association between COX-2 overexpression and angiogenesis in gastric cancer.

In conclusion, our data suggest that COX-2 overexpression leads to increased PGE2 biosynthesis and angiogenesis,which may be mechanisms underlying the contribution of COX-2 to the development of gastric cancer. Thanks to the recent advances in the development of selective COX-2 inhibitors, COX-2 may, thus, provide an attractive target for chemopreventive strategies in the treatment of gastric cancer.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

                
2

The abbreviations used are: COX, cyclooxygenase;PG, prostaglandin; PGE2, prostaglandin E2;TXB2, thromboxane B2; PGI2,prostacyclin; 6-k-PGF, 6-keto-prostaglandin F; NSAID, nonsteroidal anti-inflammatory drug.

Fig. 1.

A representative immunoblot analysis of COX-2 protein expression in the normal gastric mucosa (N) and cancer tissue (C). The cancer tissue showed intense immunoreactive bands located at Mr70,000 (70 kDa), although the paired normal gastric mucosa showed weak bands. The ratio of COX-2 C:N was more than 1.0 in 31 (74%) of 42 cases.

Fig. 1.

A representative immunoblot analysis of COX-2 protein expression in the normal gastric mucosa (N) and cancer tissue (C). The cancer tissue showed intense immunoreactive bands located at Mr70,000 (70 kDa), although the paired normal gastric mucosa showed weak bands. The ratio of COX-2 C:N was more than 1.0 in 31 (74%) of 42 cases.

Close modal
Fig. 2.

Detection of microvessels in cancer tissue by immunostaining for CD34. The microvessels were found to be most numerous at the periphery of the tumor (“hot spot”).

Fig. 2.

Detection of microvessels in cancer tissue by immunostaining for CD34. The microvessels were found to be most numerous at the periphery of the tumor (“hot spot”).

Close modal
Fig. 3.

Correlation between the microvessel density and the ratio of COX-2 C:N (A) and PGE2 levels in the cancer tissue specimens (B). The correlation coefficient between the microvessel density and the ratio of COX-2 C:N was 0.581 (P < 0.0001), whereas the coefficient between the microvessel density and PGE2 levels was −0.147(P = 0.3555).

Fig. 3.

Correlation between the microvessel density and the ratio of COX-2 C:N (A) and PGE2 levels in the cancer tissue specimens (B). The correlation coefficient between the microvessel density and the ratio of COX-2 C:N was 0.581 (P < 0.0001), whereas the coefficient between the microvessel density and PGE2 levels was −0.147(P = 0.3555).

Close modal
Table 1

The association of COX-2 expression with the histological findings

COX-2 C:NaP                  b
Histological typec   
Intestinal (n = 17) 3.7 ± 3.3 NSd 
Diffuse (n = 25) 2.9 ± 3.2  
Depth of invasione   
T1, T2 (n = 24) 2.7 ± 2.4 NS 
T3, T4 (n = 18) 4.1 ± 4.0  
Lymph node involvement   
Negative (n = 13) 1.9 ± 1.4 <0.05 
Positive (n = 29) 3.9 ± 3.7  
COX-2 C:NaP                  b
Histological typec   
Intestinal (n = 17) 3.7 ± 3.3 NSd 
Diffuse (n = 25) 2.9 ± 3.2  
Depth of invasione   
T1, T2 (n = 24) 2.7 ± 2.4 NS 
T3, T4 (n = 18) 4.1 ± 4.0  
Lymph node involvement   
Negative (n = 13) 1.9 ± 1.4 <0.05 
Positive (n = 29) 3.9 ± 3.7  
a

The values are expressed as the mean ± SD.

b

According to Welch’s t test.

c

According to Lauren’s classification.

d

NS, not significant.

e

According to the tumor-node-metastasis classification.

Table 2

The levels of PGs in human gastric cancers

PG levelsa (ng/mg protein)PGE2TXB26-k-PGF
NCNCNC
Total (n = 42) 166 ± 63 198 ±b83 54 ± 23 59 ± 26 61 ± 25 71 ± 24 
Histological typec       
Intestinal (n = 17) 152 ± 70 210 ± 104 49 ± 23 61 ± 23 57 ± 29 71 ± 25 
Diffuse (n = 25) 175 ± 58 190 ± 65 57 ± 22 57 ± 28 64 ± 22 70 ± 24 
Depth of invasiond       
T1, T2 (n = 24) 170 ± 73 211 ± 96 50 ± 23 55 ± 26 61 ± 27 69 ± 28 
T3, T4 (n = 18) 160 ± 48 180 ± 59 58 ± 23 63 ± 26 62 ± 22 72 ± 19 
Lymph node involvement       
Negative (n = 13) 176 ± 79 215 ± 67 57 ± 28 63 ± 27 66 ± 29 75 ± 25 
Positive (n = 29) 161 ± 56 190 ± 88 52 ± 20 57 ± 25 59 ± 23 69 ± 25 
COX-2 overexpressione       
Positive (n = 31) 161 ± 57 201 ±b90 55 ± 20 60 ± 26 62 ± 21 71 ± 26 
       
Negative (n = 11) 179 ± 80 189 ± 58 50 ± 31 54 ± 24 59 ± 34 68 ± 20 
PG levelsa (ng/mg protein)PGE2TXB26-k-PGF
NCNCNC
Total (n = 42) 166 ± 63 198 ±b83 54 ± 23 59 ± 26 61 ± 25 71 ± 24 
Histological typec       
Intestinal (n = 17) 152 ± 70 210 ± 104 49 ± 23 61 ± 23 57 ± 29 71 ± 25 
Diffuse (n = 25) 175 ± 58 190 ± 65 57 ± 22 57 ± 28 64 ± 22 70 ± 24 
Depth of invasiond       
T1, T2 (n = 24) 170 ± 73 211 ± 96 50 ± 23 55 ± 26 61 ± 27 69 ± 28 
T3, T4 (n = 18) 160 ± 48 180 ± 59 58 ± 23 63 ± 26 62 ± 22 72 ± 19 
Lymph node involvement       
Negative (n = 13) 176 ± 79 215 ± 67 57 ± 28 63 ± 27 66 ± 29 75 ± 25 
Positive (n = 29) 161 ± 56 190 ± 88 52 ± 20 57 ± 25 59 ± 23 69 ± 25 
COX-2 overexpressione       
Positive (n = 31) 161 ± 57 201 ±b90 55 ± 20 60 ± 26 62 ± 21 71 ± 26 
       
Negative (n = 11) 179 ± 80 189 ± 58 50 ± 31 54 ± 24 59 ± 34 68 ± 20 
a

Mean ± SD; N, normal gastric mucosa; C, cancer tissue.

b

P < 0.05, according to Welch’s t test.

c

According to Lauren’s classification.

d

According to the Tumor-Node-Metastasis classification.

e

Positive, COX-2 C:N > 1.0; Negative,COX-2 C:N < 1.0.

1
DeWitt D. L. Prostaglandin endoperoxide synthase: regulation of enzyme expression.
Biochim. Biophys. Acta
,
1083
:
121
-134,  
1991
.
2
Kujubu D. A., Reddy S. T., Fletcher B. S., Herschman H. R. Expression of the protein product of the prostaglandin synthase-2/TIS10 gene in mitogen-stimulated Swiss 3T3 cells.
J. Biol. Chem.
,
268
:
5425
-5430,  
1993
.
3
Kargman S., Charleson S., Cartwright M., Frank J., Riendeau D., Mancini J., Evans J., O’Neill G. Characterization of prostaglandin G/H synthase 1 and 2 in rat, dog, monkey, and human gastrointestinal tracts.
Gastroenterology
,
111
:
445
-454,  
1996
.
4
Vane J. Towards a better aspirin.
Nature (Lond.)
,
367
:
215
-216,  
1994
.
5
Kargman S. L., O’Neill G. P., Vickers P. J., Evans J. F., Mancini J. A., Jothy S. Expression of prostaglandin G/H synthase-1 and -2 protein in human colon cancer.
Cancer Res.
,
55
:
2556
-2559,  
1995
.
6
Kutchera W., Jones D. A., Matsunami N., Groden J., McIntyre T. M., Zimmerman G. A., White R. L., Prescott S. M. Prosta-glandin H synthase 2 is expressed abnormally in human colon cancer: evidence for a transcriptional effect.
Proc. Natl. Acad. Sci. USA
,
93
:
4816
-4820,  
1996
.
7
Eberhart C. E., Coffey R. J., Radhika A., Giardiello F. M., Ferrenbach S., DuBois R. N. Up-regulation of cyclooxygenase-2 gene expression in human colorectal adenomas and adenocarcinomas.
Gastroenterology
,
107
:
1183
-1188,  
1994
.
8
Sano H., Kawahito Y., Wilder R. L., Hashiramoto A., Mukai S., Asai K., Kimura S., Kato H., Kondo M., Hla T. Expression of cyclooxygenase-1 and -2 in human colorectal cancer.
Cancer Res.
,
55
:
3785
-3789,  
1995
.
9
Thun M. J., Namboodiri M. M., Heath C. W., Jr. Aspirin use and reduced risk of fatal colon cancer.
N. Engl. J. Med.
,
325
:
1593
-1596,  
1991
.
10
Thun M. J., Namboodiri M. M., Calle E. E., Heath C. W., Jr. Aspirin use and risk of fatal colon cancer.
Cancer Res.
,
53
:
1322
-1327,  
1993
.
11
Schreinemachers D. M., Everson R. B. Aspirin use and lung, colon, and breast cancer incidence in a prospective study.
Epidemiology
,
5
:
138
-146,  
1994
.
12
Hida T., Yatabe Y., Achiwa H., Muramatsu H., Kozaki K., Nakamura S., Ogawa M., Mitsudomi T., Sugiura T., Takahashi T. Increased expression of cyclooxygenase-2 occurs frequently in human lung cancers, specifically in adenocarcinomas.
Cancer Res.
,
58
:
3761
-3764,  
1998
.
13
Hwang D., Scollard D., Byrne J., Levine E. Expression of cyclooxygenase-1 and cyclooxygenase-2 in human breast cancer.
J. Natl. Cancer Inst.
,
90
:
455
-460,  
1998
.
14
Wilson K. T., Fu S., Ramanujam K. S., Meltzer S. J. Increased expression of inducible nitric oxide synthase and cyclooxygenase-2 in Barrett’s esophagus and associated adenocarcinomas.
Cancer Res.
,
58
:
2929
-2934,  
1998
.
15
Uefuji K., Ichikura T., Mochizuki H., Shinomiya N. Expression of cyclooxygenase-2 protein in gastric adenocarcinoma.
J. Surg. Oncol.
,
69
:
168
-172,  
1998
.
16
Ristimaki A., Honkanen N., Jankala H., Sipponen P., Harkonen M. Expression of cyclooxygenase-2 in human gastric carcinoma.
Cancer Res.
,
57
:
1276
-1280,  
1997
.
17
Rigas B., Goldman I. S., Levine L. Altered eicosanoid levels in human colon cancer.
J. Lab. Clin. Med.
,
122
:
518
-523,  
1993
.
18
Bennett A., Civier A., Hensby C. N., Melhuish P. B., Stamford I. F. Measurement of arachidonate and its metabolites extracted from human normal and malignant gastrointestinal tissues.
Gut
,
28
:
315
-318,  
1987
.
19
McLemore T. L., Hubbard W. C., Litterst C. L., Liu M. C., Miller S., McMahon N. A., Eggleston J. C., Boyd M. R. Profiles of prostaglandin biosynthesis in normal lung and tumor tissue from lung cancer patients.
Cancer Res.
,
58
:
3140
-3147,  
1998
.
20
Tsujii M., Kawano S., Tsuji S., Sawaoka H., Hori M., DuBois R. N. Cyclooxygenase regulates angiogenesis induced by colon cancer cells.
Cell
,
93
:
705
-716,  
1998
.
21
Martin L., Holcombe C., Green B., Leinster S. J., Winstanley J. Is a histological section representative of whole tumour vascularity in breast cancer?.
Br. J. Cancer
,
76
:
40
-43,  
1997
.
22
Martin L., Green B., Renshaw C., Lowe D., Rudland P., Leinster S. J., Winstanley J. Examining the technique of angiogenesis assessment in invasive breast cancer.
Br. J. Cancer
,
76
:
1046
-1054,  
1997
.
23
Goodwin J. S., Bankhurst A. D., Messner R. P. Suppression of human T-cell mitogenesis by prostaglandin; existence of a prosta-glandin-producing suppressor cell.
J. Exp. Med.
,
146
:
1719
-1734,  
1977
.
24
Bockman, R. S. PGE inhibition of T-lymphocyte colony formation. J. Clin. Invest., 64: 812–821, 1797.
25
Brunda M. J., Herberman R. B., Holden H. T. Inhibition of murine natural killer cell activity by prostaglandins.
J. Immunol.
,
124
:
2682
-2688,  
1980
.
26
Baich C. M., Doghert P. A., Cloud G. A., Tilden A. B. Prostaglandin E2-mediated suppression of cellular immunity in colon cancer patients.
Surgery
,
95
:
71
-77,  
1984
.
27
Honn K. V., Meyer J. Thromboxanes and prostacyclin: positive and negative modulators of tumor growth.
Biochem. Biophys. Res. Commun.
,
102
:
1122
-1129,  
1981
.
28
Hamberg M., Samuelsson B. Prostaglandin endoperoxides: novel transformations of arachidonic acid in human platelets.
Proc. Natl. Acad. Sci. USA
,
71
:
3400
-3404,  
1974
.
29
Honn K. V., Tang D. G., Chen Y. Q. Platelets and cancer metastasis: more than an epiphenomenon.
Semin. Thromb. Hemostasis
,
18
:
390
-413,  
1992
.
30
Tsujii M., Kawano S., DuBois R. N. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential.
Proc. Natl. Acad. Sci. USA
,
94
:
3336
-3340,  
1997
.
31
Falkman J. What is the evidence that tumors are angiogenesis dependent?.
J. Natl. Cancer Inst.
,
82
:
4
-6,  
1990
.
32
Bouck N. Tumor angiogenesis: the role of oncogenes and tumor suppressor genes.
Cancer Cells (Cold Spring Harbor)
,
2
:
179
-185,  
1990
.