| Japanese Journal of Clinical Oncology | Pages |
Increased Expression of Cyclooxygenase-2 to -1 in Human Colorectal Cancers and Adenomas, but not in Hyperplastic Polyps
Introduction
Methods
Subjects
DNA Extraction, RNA Preparation and cDNA Synthesis
Design and Synthesis of Fluorescence-labeled Primers
PCR Procedures
SSCP Analysis
Immunohistochemistry
Results
Experimental Conditions for SSCP Analysis
The Proportion of COX-2 mRNA in Colorectal Tissues, Cancer, Adenoma and Hyperplastic Polyps
Discussion
Acknowledgments
References
Increased Expression of Cyclooxygenase-2 to -1 in Human Colorectal Cancers and Adenomas, but not in Hyperplastic Polyps
Results: The present F-SSCP analysis was a simple and powerful method for quantitative determination of the proportions of COX-2 mRNA. The proportion of COX-2 mRNA was higher in cancer tissues than in accompanying normal mucosa in 46 of the 50 cancers. There was no significant correlation between the increase of the COX-2 proportion and tumor location or stages. The enhanced COX-2 expression was also observed in colorectal adenomas. On the other hand, the proportion of COX-2 mRNA in hyperplastic polyps was not significantly different from that in normal mucosa.
Conclusions: The proportion of COX-2 to COX-1 expression was elevated in most human colorectal cancers and adenomas, but not in hyperplastic polyps. Therefore, the increased proportion of COX-2 expression might be an early event in the carcinogenesis of colorectal cancer.
INTRODUCTION
Recent studies have shown that non-steroidal anti-inflammatory drugs (NSAIDs), which inhibit cyclooxygenase (COX: prostaglandin endoperoxide synthase; EC 1.14.99.1) and the synthesis of prostaglandins (PGs), can reduce the formation of colon cancers in experimental animals given carcinogens (1). Epidemiological analyses have indicated that administration of NSAIDs can reduce the incidence of colorectal cancer in humans (2-4). Mammalian cells contain two related, but unique, isoforms of COX, referred to as COX-1 and COX-2. These two proteins are encoded by separate genes (5,6). The most dramatic difference between these two isozymes is observed in their patterns of expression. COX-1 is present in most tissues and is involved in the physiological production of PGs for maintaining normal homeostasis, whereas COX-2, which is induced by mitogens, cytokines and growth factors, is primarily responsible for PGs produced in inflammatory sites (7-9). Eberhart et al. (10) have shown that COX-2, but not COX-1, mRNA expression is markedly elevated in the tissues of most human colorectal cancers and some colorectal adenomas compared with accompanying normal mucosa. More recent studies suggest that COX-2 is related to colon carcinogenesis and may be the target for the chemopreventive effect of NSAIDs. Elevated levels of COX-2 protein and mRNA, but not those of COX-1, are found in chemically introduced rat colon carcinoma tissues (11,12) and in human colon carcinoma (12-15). Furthermore, genetic disruption of the COX-2 gene or treatment with a COX-2 specific drug suppresses the polyp formation in mice for familial adenomatous polyposis and marked increases in COX-2 enzyme concentrations are found in early polyps in these animals (16).
Previously, we reported the combined use of an automated fluorescence-based DNA sequencer and a data processing computer, allowing the exact quantification of polymorphic DNA sequences and the combination of reverse transcribed polymerase chain reaction (RT-PCR) and single-strand conformation polymorphism (SSCP) analysis, for the quantitative detection of expression levels of genes encoding homologous sequences (17,18). The purpose of the present study was to establish a simple procedure to analyze proportions of COX-2 and COX-1 mRNAs using RT-PCR-SSCP, thus permitting application to small amounts of biopsied specimens or surgical resected specimens (colorectal cancer, adenoma and polyps). The principle of the present investigation is based on the condition that COX-1 is constantly expressed in most tissues and that the proportion of COX-2 to total COX mRNA could reflect the amount of COX-2 mRNA.
METHODS
Subjects
Colorectal tissues were obtained from the National Cancer Center Hospital, cancer and normal accompanying mucosa resected at surgery and hyperplastic polyps, adenoma and corresponding normal mucosa at endoscopic biopsy. The tissues were stored in liquid nitrogen or at -80°C until use.
DNA Extraction, RNA Preparation and cDNA Synthesis
DNA was extracted from tissue specimens by a method described previously (19). Total RNA was prepared from various cancer cell lines and colorectal tissues using a standard guanidium thiocyanate-phenol-chloroform-isoamyl alcohol extraction method (20). Reverse transcriptase reactions (20 µl) were carried out with [sim]0.5-1.0 µg of total RNA as templates and random hexamer as primers, using SuperScript RNase free reverse transcriptase (GIBCO BRL, Gaithersburg, MD).
Design and Synthesis of Fluorescence-labeled Primers
Oligonucleotide primers were prepared using an Oligo 1000 DNA synthesizer (Beckman Instruments, Fullerton, CA). In order to coamplify both COX-1 and COX-2 cDNAs, we designed a set of PCR primer pairs that anneal to the conserved region of COX cDNA sequences, i.e. in exons 5, 6 and 7 (Fig. 1) (21-23). Both primers contained several mismatches to the target sequences of either COX-2 or COX-1 cDNA. The nucleotide sequences of the primers were 5[prime]-AAG-TTC-ATA-CCT-GAT-CCC-CA-3[prime] (forward) and 5[prime]-GGA-GGA-TAC-ATT-TCT-CCA-TC-3[prime] (reverse). The primers yielded a product of 236 bp from both COX-2 and COX-1 cDNAs. The 5[prime]-terminus of the forward primer was labeled with indodicarbocyanine (Cy5 amidite) fluorescent dye, using ALFred (Cy5 amidite) reagent (Pharmacia, Uppsala, Sweden).
Figure 1. Homology alignment of the nucleotide and amino acid sequences highly conserved in human COX-1 and COX-2 and PCR primer sequences for COX cDNA. Part of the nucleotide and amino acid sequences of COX-1 and those of COX-2 that differ from those of COX-1 are indicated. Identical nucleotides are indicated by hyphens and primer sequences are denoted by bold characters. We also designed specific primers for COX-2 and COX-1 to make cDNA fragments containing sequences of either COX-2 or COX-1. The nucleotide sequences of the primers were 5[prime]-CCA-GTA-TAA-GTG-CGA-TTG-TAC-C-3[prime] (forward; 211-232) and 5[prime]-CCG-TAG-ATG-CTC-AGG-GAC-TTG-A-3[prime] (reverse; 907-886) for COX-2 and 5[prime]-TCC-ATG-CCA-GCA-CCA-GGG-CAT-C-3[prime] (forward; 114-135) and 5[prime]-TGG-CTC-TGG-GGC-GGG-ATG-CCT-C-3[prime] (reverse; 848-827) for COX-1.
PCR Procedures
The cDNAs were amplified by PCR using a Cy5 labeled primer. The reaction mixture (25 µl) contained 10 mmol/l Tris-HCl (pH 8.3), 50 mmol/l KCl, 1.5 mmol/l MgCl2, 1 g/l gelatin and 0.2 mmol/l each of four dNTPs, oligonucleotide primers (6 pmol each), 0.6 units of AmpliTaq polymerase (Perkin-Elmer-Cetus, Branchburg, NJ) and 0.5 µl of the cDNA prepared above. After the first denaturation step at 94°C for 3 min, 45 cycles of reaction were performed at 94°C for 30 s, at 53°C for 30 s and at 72°C for 1 min, with a final extension at 72°C for 7 min. The yields of amplified DNA fragments were determined by electrophoresis in 8% polyacrylamide gels and visualized by ethidium bromide staining. A 5 µl volume of the reaction mixture was mixed with 0.5 units of Klenow fragment (Takara Shuzo, Shiga, Japan) and incubated at 37°C for 1 h in order to generate blunt-end (16,17).
SSCP Analysis
An ALFred DNA sequencer (Pharmacia) equipped with a short gel plate (200 mm height × 345 mm width × 0.35 mm thickness) was used for SSCP analysis. A 1 µl volume of the amplified products after treatment with Klenow fragment was mixed with 10 µl of loading solution containing 90% deionized formamide, 20 mmol/l EDTA and 0.05% bromophenol blue. After denaturation at 80°C for 5 min, 1 µl of the mixture was applied to a 10% polyacrylamide gel (the ratio of acrylamide and bisacrylamide was 30:1) containing Tris-glycine buffer (25 mmol/l Tris, 192 mmol/l glycine). Electrophoresis was performed at 20 W for 5 h at 20°C, using a running buffer consisting of 25 mmol/l Tris and 192 mmol/l glycine. Electrophoretic profiles were analyzed by the software Fragment Manager (Pharmacia). The proportion of COX-2 cDNA to total COX cDNA was calculated as COX-2 versus COX-2 plus COX-1, from the area of the signals for COX-2 and COX-1.
Immunohistochemistry
Immunohistochemical staining was performed with the Vectastain avidin-biotin peroxidase complex kit (Vector, Burlingame, CA) following previous reports (14,21,24). The antisera used were raised against human COX-1 and COX-2 polypeptide (14,24). Colon tissues were preserved in 10% formalin and the specimens were embedded in paraffin, serially sectioned onto microscope slides at a thickness of 4 µm. The slides were stained with antibodies against COX-2 and COX-1 and normal rabbit serum as a negative control at dilution of 1:100. In the final step, color was developed with a solution containing 0.02% peroxide, tetrahydrochloride, 0.04% nickel chloride and 0.01% hydrogen peroxide in 0.05 M Tris-HCl (pH 7.2) for 2-5 min. The sections were counterstained with 0.5% light green (Sigma, St Louis, MO). Specificity was determined by preadsorption of anti-COX-1 or -2 antibody with the COX-1 or -2 synthetic polypeptide (1 mg/ml) before staining. For each tissue specimen, the extent and intensity of staining with COX-1 and -2 antibodies were graded on a scale of 0 to 4+, as described previously (24).
RESULTS
Experimental Conditions for SSCP Analysis
The forward primer anneals COX-1 and COX-2 sequences with one mismatch and the reverse primer anneals them with COX-1 and COX-2 sequences with two and one mismatches, respectively. The COX-1 or COX-2 could be amplified by the primers described above from solutions containing COX-1 or COX-2 cDNA, respectively. In PCR amplification, a combination of a Cy5 labeled forward primer and a non-labeled reverse primer provided a good separation of DNA fragments from COX-2 and COX-1 cDNAs by the SSCP analysis as used in the present experiments (Fig. 2). The peaks of COX-1 and COX-2 were electrophoresed with similar migrations in both normal and cancer tissues. The relative proportion of COX-2 to COX-2 plus COX-1 was calculated from the densitometric curves.
Figure 2. RT-PCR-SSCP profiles of COX-2 and COX-1 in human surgical specimens. The electropherograms were analyzed by the software Fragment Manager (Pharmacia). The peaks of the COX-2 and COX-1 were identified by electrophoretic migration of the amplified products using specific primers for COX-2 and COX-1, respectively. A proportion of COX-2 mDNA to total COX (COX-2 plus COX-1) was calculated from the areas of the signals for COX-2 and COX-1. COX-2 and COX-1 cDNAs (each [sim]10 pg/µl) amplified by using the specific primers for them were mixed in different ratios (1:9, 2:8, 3:7, 4:6, 5:5, 6:4, 7:3, 8:2, 9:1). The mixtures were subjected to PCR followed by SSCP analysis. Determination of the proportion of COX-2 in the mixture confirmed good linearity, as in previous reports using lactate dehydrogenase isozymes (Fig. 3). Figure 3. Quantitative analysis of the proportions of COX-2 cDNA to total COX cDNA. COX-2 and COX-1 cDNA samples which were almost equivalent in molar concentrations were mixed in different ratios (9:1, 8:2, 7:3, 6:4, 5:5, 4:6, 3:7, 2:8 and 1:9). The DNA fragments were amplified by PCR and analyzed by SSCP (in duplicate) to detect the proportions of COX-2 mRNA. The areas of the signals analyzed by SSCP were calculated by the software Fragment Manager and the results were plotted against the COX-2 to COX-1 ratios in the mixtures.
Table 1.
| Case No. | Age (yr) | Gender | Location | Histological differentiation | Normal | Cancer | ||||
| COX-1* | COX-2* | COX-2 (%)[dagger] | COX-1* | COX-2* | COX-2 (%)[dagger] | |||||
| 7 | 53 | M | Sigmoid colon | Moderately | 0 | 0 | 36.7 | 0 | 1.5+ | 99.9 |
| 10 | 50 | F | Rectum | Well | 0 | 0 | 39.0 | 0 | 1.5-2+ | 90.5 |
| 21 | 73 | F | Ascending colon | Poorly | 0 | 0 | 35.9 | 0 | 0-0.5+ | 97.9 |
| 23 | 51 | M | Rectum | Moderately | 0 | 0 | 27.5 | 0 | 0.5-1+ | 99.9 |
| 26 | 49 | M | Sigmoid colon | Well | 0 | 0 | 74.1 | 0 | 1.5-2+ | 99.9 |
| 28 | 62 | F | Rectum | Well | 0 | 0 | 20.0 | 0-0.5+ | 3.5-4+ | 97.0 |
| 71 | 87 | M | Ascending colon | Moderately | 0 | 0 | 8.3 | 0-0.5+ | 4+ | 97.3 |
| 88 | 60 | F | Ascending colon | Well | 0 | 0 | 49.5 | 0.5+ | 4+ | 74.2 |
| 95 | 21 | M | Sigmoid colon | Moderately | 0 | 0 | 1.0 | 0.5+ | 2.5-3+ | 96.2 |
The Proportion of COX-2 mRNA in Colorectal Tissues, Cancer, Adenoma and Hyperplastic Polyps
A total of 50 cases of paired adenocarcinoma and normal mucosa obtained from surgical materials were evaluated (Fig. 4). The proportion of COX-2 mRNA to total COX mRNA was higher in cancer tissues than in normal mucosa in 46 of the 50 cases. Ten pairs of colorectal adenoma, nine pairs of colorectal hyperplastic polyps and corresponding normal mucosa were also evaluated. The proportion of COX-2 mRNA was higher in adenoma than in normal mucosa in nine of the 10 cases, whereas in hyperplastic polyps only three of the nine cases showed slightly higher proportions of COX-2 mRNA than normal mucosa. By a paired t-test, the mean of the proportion in cancer tissues and adenoma is significantly higher (p < 0.01) than that in normal mucosa. On the other hand, that in hyperplastic polyps is not significantly different from that in normal mucosa (p < 0.05).
Figure 4. Proportion of the COX-2 mRNA to total COX mRNA in colorectal tissues. Colorectal cancer (50 cases), adenoma (10) and hyperplastic polyps (9) and the corresponding normal mucosa were subjected to the present analysis for determining the proportion of the COX-2 mRNA. In several cases, colonic tissues from colorectal cancers and normal tissues were evaluated for immunoreactive COX-1 and -2 with specific antisera. The intensity of immunostaining was evaluated as shown in Table 1. In tissues of colorectal cancers, marked expression of immunoreactive COX-2 was shown whereas staining of immunoreactive COX-1 was very weak.
The proportion of COX-2 mRNA to total COX was classified according to degree of differentiation, location of the cancer and clinical stage (Fig. 5). No conditions had a significant effect on the proportion of COX-2 (p < 0.05) (by the Mann-Whitney test).
Figure 5. Influence of degree of differentiation, location and Dukes' stage on the proportion of COX-2 mRNA to total COX mRNA. The proportion of COX-2 mRNA in each cancer tissue was plotted according to degree of differentiation, location of the cancer and clinical stage. Poorly differentiated cancers showed higher proportions of the COX-2 mRNA than well and moderately differentiated cancers. Recent clinical and epidemiological observations have suggested that NSAIDs decrease the incidence of colorectal cancer (2-4,25). In patients with familial adenomatous polyposis, administration of sulindac resulted in a striking reduction in adenoma size and number (26). Moreover, in animal models of colon carcinogenesis, COX inhibitors including indomethacin, sulindac and piroxicam exhibit chemopreventive effects as judged by reductions in the number of tumors per animal. COX catalyzes the committed step in the formation of prostaglandins and thromboxane from arachidonic acid and also catalyzes the oxidation of a wide range of biological materials including several classes of chemical carcinogens (27,28). The synthesis of COX, especially COX-2, is stimulated by growth factors or cytokines such as interleukin 1, tumor necrosis factor, epidermal growth factor and platelet-derived growth factor (29,30) and prostaglandins also seem to modulate cellular proliferation in a variety of cell types (31,32). Although the precise mechanism by which NSAIDs inhibit colon carcinogenesis is unknown, prostaglandins and COX enzymes may be involved in the process of carcinogenesis because a major action of NSAIDs is the inhibition of COX. Recently, Tsujii and DuBois (33) permanently transfected rat intestinal epithelial cells with COX-2. The transfected epithelial cells expressed elevated COX-2 protein levels and demonstrated increased adhesion to extracellular matrix (ECM) proteins. The epithelial cells were also resistant to butyrate-induced apoptosis, had elevated bcl2 protein expression and reduced transforming growth factor [beta]2 receptor levels. These phenotypic changes involving both increased adhesion to ECM and inhibition of apoptosis were reversed by sulindac (a COX inhibitor). The results might indicate that overexpression of COX-2 leads to phenotypic changes in intestinal epithelial cells that could enhance their tumorigenic potential, via interference with cell adhesion and apoptosis mechanisms. Some reports have indicated that human colon cancer tissues contain larger amounts of COX-2 protein or mRNA than surrounding normal mucosa (10-16). In a report by Eberhart et al. (10), the up-regulation of COX-2 did not correlate with the stage or degree of differentiation of the carcinoma. COX-2 but not COX-1 gene expression was markedly elevated in human colorectal cancers and carcinogen-induced rat colonic tumors (11,12). COX-2 was transcribed abnormally in human colon cancers, resulting in the abnormal expression of COX-2 (15). Marked increases in COX-2 enzyme concentrations have been found in early polyps in mice with defects in the APC gene (16). In the present study, the increased proportion of COX-2 expression was observed in most colorectal cancers and adenomas but not in hyperplastic polyps. The results suggested that the up-regulation of COX-2 expression might be an early event in the disease process. We also showed that the increased proportion of COX-2 expression was consistently observed in any stage, tumor location or degree of differentiation. In mammalian cells COX-1 and COX-2 proteins are encoded by two separate genes (5,6). The deduced amino acid sequences of COX-2 exhibited about 60% similarities to the sequences of COX-1 proteins and 81-88% similarities among the COX-2 enzymes of different species. The amino acid residues established to be important in catalysis by COX-1 are conserved in COX-2 (5,6). In particular, since human COX-2 exons 4-10 show over 70% identity to COX-1 exons at the amino acids level, we selected PCR primers in exons 5, 6 and 7. COX-2 cDNA, which encodes a polypeptide of 604 amino acids, is about 58% identical with the COX-1 cDNA in the nucleotide sequence level (21-23). The present procedure is essentially based on the assumption that COX-1 expression is constant in tissues. Reportedly, COX-1 is present in most tissues and mostly constant in any physiological conditions (7-9). This means that our procedure detects only relative ratios of COX-2 expression against COX-1 expression. However, COX-1 might be an adequate control to detect the COX-2 expression level. Immunohistochemical staining could not detect COX-1 in some cases and was not consistent with the result of RT-PCR-SSCP. It is less efficient and has lower sensitivity. It is also subjective and difficult to make quantitative. The slight inconsistency between the proportion of COX-2 estimated by RT-PCR-SSCP and COX-2 intensity of immunostaining might be due to the above reasons. The present procedure for determining relative ratios of COX-2 expression has the following advantages. Template concentrations and PCR cycles do not affect the proportions of COX-2 in the product if the amplified products have sufficient fluorescence signals, as was shown in our previous report (18). It has good reproducibility and linearity. No specific additional steps are required after PCR except for Klenow treatment and electrophoresis on a DNA sequencer. The method also lends itself to the simultaneous examination of multiple samples. The main advantages are its simplicity and relatively high sensitivity. Current PCR-based techniques can be applied to biopsied specimens of colorectal polyps and cancer tissues in order to investigate some properties of cancer and the preventive effects of NSAIDs on carcinogenesis. Our subsequent efforts will be directed towards a comprehensive analysis of COX-2 expression and clinical time course, including carcinogenesis, in human colorectal cancers, adenomas and polyps before and after administration of NSAIDs such as COX-2 specific inhibitor or non-specific COX inhibitor. This work was supported in part by Grants-in-Aid for Cancer Research and for the 2nd Term Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health and Welfare, Japan, and grants from the Public Trust Haraguchi Memorial Cancer Research Fund, Mitsui Life Social Welfare Foundation and Kurozumi Medical Foundation. We thank Dr M. Ochiai and Ms Yamauchi of the Pathology Division of the National Cancer Center for technical assistance and valuable suggestions.
DISCUSSION
Acknowledgments
References
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 8 Aug 1998
Copyright©Japanese Journal of Clinical Oncology, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
D. Mazhar, R. Gillmore, and J. Waxman
COX and cancer
QJM,
October 1, 2005;
98(10):
711 - 718.
[Full Text]
[PDF]
G. J. Kelloff, R. L. Schilsky, D. S. Alberts, R. W. Day, K. Z. Guyton, H. L. Pearce, J. C. Peck, R. Phillips, and C. C. Sigman
Colorectal Adenomas: A Prototype for the Use of Surrogate End Points in the Development of Cancer Prevention Drugs
Clin. Cancer Res.,
June 1, 2004;
10(11):
3908 - 3918.
[Full Text]
[PDF]
T. Sato, K. Yoshinaga, S. Okabe, T. Okawa, T. Higuchi, M. Enomoto, T. Takizawa, and K. Sugihara
Cyclooxygenase-2 Expression and its Relationship with Proliferation of Colorectal Adenomas
Jpn. J. Clin. Oncol.,
December 1, 2003;
33(12):
631 - 635.
[Abstract]
[Full Text]
[PDF]
G. Davies, L.-A. Martin, N. Sacks, and M. Dowsett
Cyclooxygenase-2 (COX-2), aromatase and breast cancer: a possible role for COX-2 inhibitors in breast cancer chemoprevention
Ann. Onc.,
May 1, 2002;
13(5):
669 - 678.
[Abstract]
[Full Text]
[PDF]
M HULL, M LANGMAN;, J DIMBERG, and P SODERKVIST
Differential expression of cyclooxygenase 2 in human colorectal cancer Reply
Gut,
July 1, 2000;
47(1):
154 - 154.
[Full Text]
[PDF]
K. S. Chapple, E. J. Cartwright, G. Hawcroft, A. Tisbury, C. Bonifer, N. Scott, A. C. J. Windsor, P. J. Guillou, A. F. Markham, P. L. Coletta, et al.
Localization of Cyclooxygenase-2 in Human Sporadic Colorectal Adenomas
Am. J. Pathol.,
February 1, 2000;
156(2):
545 - 553.
[Abstract]
[Full Text]
[PDF]
S. H. Erdman, N. A. Ignatenko, M. B. Powell, K. A. Blohm-Mangone, H. Holubec, J. M. Guillen-Rodriguez, and E. W. Gerner
APC-dependent changes in expression of genes influencing polyamine metabolism, and consequences for gastrointestinal carcinogenesis, in the Min mouse
Carcinogenesis,
September 1, 1999;
20(9):
1709 - 1713.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (43)
Request Permissions
Google Scholar
Articles by Maekawa, M
Articles by Sekiya, T
Search for Related Content
PubMed
PubMed Citation
Articles by Maekawa, M
Articles by Sekiya, T
Social Bookmarking
What's this?