A selective cyclooxygenase 2 inhibitor suppresses the growth of H-ras-transformed rat intestinal epithelial cells.

BACKGROUND & AIMS: Constitutive expression of cyclooxygenase 2 (COX-2) has been found in 85% of colorectal cancers. Ras mutations are found in 50% of colorectal adenocarcinomas. The aim of this study was to determine the role of COX-2 in ras-induced transformation in rat intestinal epithelial (RIE) cells. METHODS: Cell growth was determined by cell counts. The expression of COX-2 was examined by Northern and Western analyses. For tumorigenicity assays, cells were inoculated into dorsal subcutaneous tissue of athymic nude mice. DNA-fragmentation assays were performed to detect apoptosis. RESULTS: The expression of COX-2 was increased in RIE-Ras cells at both messenger RNA (9-fold) and protein (12-fold) levels. Prostaglandin I2 levels were elevated 2.15-fold in RIE-Ras cells. Serum deprivation further increased COX-2 expression 3.8-fold in RIE-Ras cells. Treatment with a selective COX-2 antagonist (SC58125) inhibited the growth of RIE-Ras cells through inhibition of cell proliferation and by induction of apoptosis. SC-58125 treatment reduced the colony formation in Matrigel by 83.0%. Intraperitoneal administration of SC-58125 suppressed RIE-Ras tumor growth in nude mice by 60.3% in 4 weeks. SC-58125 treatment also induced apoptosis in RIE-Ras cells as indicated by increased DNA fragmentation. CONCLUSIONS: Overexpression of COX-2 may contribute to tumorigenicity of ras-transformed intestinal epithelial cells. Selective inhibition of COX-2 activity inhibits growth of ras-transformed intestinal epithelial cells and induces apoptosis.

Cyclooxygenase-2 induction and transforming growth factor beta growth inhibition in rat intestinal epithelial cells.

Rat intestinal epithelial cells (RIE-1) permanently transfected with the prostaglandin endoperoxide synthase 2 (also referred to as cyclooxygenase-2; COX-2) gene exhibit decreased cyclin D1 levels, decreased cdk4-associated kinase activity, and delayed G1 cell cycle progression, which represents a phenotype similar to that which follows transforming growth factor beta (TGF-beta) treatment. In the current study, we have found that addition of TGF-beta 1 to the parental RIE-1 cells (designated RIE-P) caused a rapid induction of COX-2 mRNA and protein. COX-2 protein levels progressively increased and reached peak levels 6 h after TGF-beta 1 addition. Cyclin D1 was decreased by 74% at 6 h and was undetectable 24 h after addition of TGF-beta 1. In RIE cells transfected with the COX-2 antisense expression vector (RIE-AS cells), TGF-beta 1 induction of COX-2 protein was reduced greater than 90%. Addition of TGF-beta 1 did not reduce the abundant cyclin D1 protein expression in the RIE-AS cells, unlike the effect in RIE-P cells. TGF-beta 1 treatment reduced peak [3H]thymidine incorporation by 60% and delayed G1/S-phase transition by at least 4 h in the RIE-P cells. In contrast, S-phase entry occurred at 16 h in RIE-AS cells and was not altered by TGF-beta 1 treatment. Restoration of cyclin D1 expression by transfection of the cyclin D1 cDNA under transcriptional control of the cytomegalovirus promoter/enhancer in the COX-2-overexpressing (RIE-S) cells decreased the time required for S-phase entry by at least 4 h and increased the peak level of [3H]thymidine incorporation. Taken together, the results demonstrate that TGF-beta 1 strongly induces COX-2 at both the mRNA and protein levels and suggest that this induction of COX-2 is involved in the down-regulation of cyclin D1 and inhibition of cell growth caused by TGF-beta 1 in rat intestinal epithelial cells.