Heterogeneous distribution of COX-2 over-expression in human colon carcinoma cells

Nuclear factor-kappaB mediates over-expression of cyclooxygenase-2 during activation of RAW 264.7 macrophages in selenium deficiency

Selenium (Se) is an essential micronutrient for all mammalian species and is associated with a variety of physiological functions, notably immune system, in the form of selenoproteins. Inadequate Se nutrition has been linked to various diseases, including rheumatoid arthritis, cardiomyopathy, and cancer. Important to this discussion is that cyclooxygenase-2 (COX-2) is over-expressed in all the aforesaid pathologies; however, a casual relationship between Se status and COX-2 expression remains to be established. The present study is based on the hypothesis that oxidant stress, a consequence of Se deficiency, lowers the activation potential of the redox-sensitive transcription factor, NF-kappaB, and that the activated NF-kappaB is required for the altered expression of COX-2. To test this hypothesis, we have investigated the relationship between Se status and COX-2 expression in response to LPS stimulation in RAW 264.7, a macrophage-like cell line. In Se-deficient cells, the Se-dependent glutathione peroxidase activity (Se-GPx), a measure of Se status, was markedly reduced and the overall oxidative stress was significantly higher than Se-supplemented cells. Upon lipopolysaccharide (LPS) stimulation, we found 2-3-folds higher COX-2 protein expression as well as higher PGE2 levels in Se-deficient cells than Se-supplemented cells. In comparison, COX-1 protein expression was not affected by either LPS stimulation or Se status. Following LPS stimulation, the nuclear localization of NF-kappaB was significantly increased in Se-deficient macrophages, thereby leading to increased expression of COX-2. This is the first report demonstrating an inverse relationship between Se status and the expression of COX-2.

Increased expression of cyclooxygenase-2 protein in 4-nitroquinoline-1-oxide-induced rat tongue carcinomas and chemopreventive efficacy of a specific inhibitor, nimesulide

Expression of cyclooxygenase (COX)-2 protein in 4-nitroquinoline-1-oxide (4-NQO)-induced rat tongue lesions and the postinitiation chemopreventive potential of a selective COX-2 inhibitor, nimesulide (NIM), were examined in Fischer 344 male rats. NIM was administered in the diet at doses of 150, 300, and 600 ppm for 14 weeks after treatment with 25-35 ppm 4-NQO in the drinking water for 12 weeks. Western blot analysis revealed COX-2 protein to be barely expressed in the normal tongue epithelia, whereas it was increased approximately 6-fold in squamous cell carcinomas (SCCs). Immunohistochemically, COX-2 protein was diffusely present in SCCs and dysplasia but expressed only in basal cells in hyperplasia and papillomas. In basal cells of normal epithelia, it was also occasionally weakly stained. NIM dose-dependently decreased at doses of 150 and 300 ppm, the incidences of SCCs to 4 of 12 (33.3%) and 1 of 13 (7.7%) and their multiplicity to 0.33+/-0.49 and 0.08+/-0.28 per rat, respectively, as compared with 4-NQO alone group values of 9 of 11 (81.8%) and 1.00+/-0.77. A lesser decrease was observed with 600 ppm, the values being 5 of 12 (41.7%) and 0.50+/-0.67. NIM did not significantly affect the development of hyperplasias, dysplasias, and papillomas. These results clearly indicate chemopreventive potential of a selective COX-2 inhibitor against the postinitiation development of SCCs in rat tongue carcinogenesis.

COX-2 is expressed in human pulmonary, colonic, and mammary tumors

BACKGROUND

The cyclooxygenase (COX) enzyme catalyzes the formation of prostaglandins, which can affect cell proliferation and alter the response of the immune system to malignant cells. The inducible form of COX, COX-2, has been shown to be important in carcinogenesis.

METHODS

The authors studied COX-1 and -2 expression in 20 tumors of the lung, colon, and breast (60 total) by using commercially available monoclonal and polyclonal antibodies on formalin fixed, paraffin embedded tissue. Our evaluation also included seven carcinoma-associated colonic adenomas and 10 mammary ductal carcinomas in situ (DCIS). Quantitation of immunoreactivity was accomplished using an immunohistochemical scoring system that approximates the use of image analysis-based systems.

RESULTS

Ninety percent of lung tumors (squamous cell carcinomas and adenocarcinomas), 71% of colon adenocarcinomas and 56% of breast tumors (DCIS and infiltrating ductal and lobular carcinomas) expressed COX-2 at a moderate to strong level, which was significantly different from the negligible expression in distant nonneoplastic epithelium (controls; P < 0.0001). Poorly differentiated histologic features were correlated with low COX-2 expression overall, especially in colon carcinomas. Among breast carcinomas, DCIS was more likely to express COX-2 than invasive carcinomas. Adenomatous colonic epithelium showed moderate COX-2 expression, as did adjacent nonneoplastic epithelium. COX-1 immunoreactivity was essentially weak to moderate in all tissues evaluated.

CONCLUSIONS

COX-2 expression is upregulated in well and moderately differentiated carcinomas of the lung, colon, and breast whereas COX-1 appears to be constitutively expressed at low levels. A possible COX-2 paracrine effect is suggested by moderate immunoreactivity in adjacent nonneoplastic epithelium.

Biochemistry of cyclooxygenase (COX)-2 inhibitors and molecular pathology of COX-2 in neoplasia

Several types of human tumors overexpress cyclooxygenase (COX) -2 but not COX-1, and gene knockout transfection experiments demonstrate a central role of COX-2 in experimental tumorigenesis. COX-2 produces prostaglandins that inhibit apoptosis and stimulate angiogenesis and invasiveness. Selective COX-2 inhibitors reduce prostaglandin synthesis, restore apoptosis, and inhibit cancer cell proliferation. In animal studies they limit carcinogen-induced tumorigenesis. In contrast, aspirin-like nonselective NSAIDs such as sulindac and indomethacin inhibit not only the enzymatic action of the highly inducible, proinflammatory COX-2 but the constitutively expressed, cytoprotective COX-1 as well. Consequently, nonselective NSAIDs can cause platelet dysfunction, gastrointestinal ulceration, and kidney damage. For that reason, selective inhibition of COX-2 to treat neoplastic proliferation is preferable to nonselective inhibition. Selective COX-2 inhibitors, such as meloxicam, celecoxib (SC-58635), and rofecoxib (MK-0966), are NSAIDs that have been modified chemically to preferentially inhibit COX-2 but not COX-1. For instance, meloxicam inhibits the growth of cultured colon cancer cells (HCA-7 and Moser-S) that express COX-2 but has no effect on HCT-116 tumor cells that do not express COX-2. NS-398 induces apoptosis in COX-2 expressing LNCaP prostate cancer cells and, surprisingly, in colon cancer S/KS cells that does not express COX-2. This effect may due to induction of apoptosis through uncoupling of oxidative phosphorylation and down-regulation of Bcl-2, as has been demonstrated for some nonselective NSAIDs, for instance, flurbiprofen. COX-2 mRNA and COX-2 protein is constitutively expressed in the kidney, brain, spinal cord, and ductus deferens, and in the uterus during implantation. In addition, COX-2 is constitutively and dominantly expressed in the pancreatic islet cells. These findings might somewhat limit the use of presently available selective COX-2 inhibitors in cancer prevention but will probably not deter their successful application for the treatment of human cancers.