Involvement of cyclooxygenase-2 in rat models of conjunctivitis

Purpose. Using two animal models to determine which isoform of cyclooxygenase (COX), constitutive COX-1 or inducible COX-2, is involved in the progression of anterior ocular inflammation. Methods. Lambda-carrageenan (500 μg/eye) or bacterial lipopolysaccharide (LPS; 3 μg/eye) was injected into rat conjunctiva to induce conjunctivitis. Vascular permeability in inflamed conjunctiva was measured by uptake of systemic Evans blue. Changes in mRNA for COX-1 and COX-2 in conjunctiva were detected by RT-PCR. Changes in COX-2 protein were detected by immunoblotting after immunoprecipitation. To assess involvement of COX-2 in carrageenan- and LPS-induced conjunctivitis, NS-398 (a selective COX-2 inhibitor) or indomethacin (non-selective COX inhibitor) was topically administrated at 15 and 30 minutes before inflammatory stimulator-injection. Results. In the carrageenan-injected model, the dye content of conjunctiva (12.4 ± 2.8 mg/eye) was significantly increased 4 hours after injection compared to saline-injected control rats (3.7 ± 1.1 mg/eye). mRNA for COX-2 was significantly increased by 2 hours and gradually increased until 24 hours; COX-1 mRNA did not show major changes until 24 hours after injection. COX-2 protein was markedly elevated 4 hours after injection of carrageenan. COX-2 protein levels were well correlated with increased mRNA levels. In the LPS-injected model, the dye content of conjunctiva (5.8 ± 1.2 mg/eye) was significantly increased 4 hours after injection compared to saline-injected control rats (3.1 ± 0.6 mg/eye). Expression of COX-2 mRNA was increased 1 hour after injection, peaked at 2 hours, and decreased at 4 hours. mRNA for COX-1 did not change by 24 hours. COX-2 protein increased 2 hours after injection of LPS. COX-2 protein levels were well correlated with increased mRNA. Topical administration of 1% NS-398 exhibited strong inhibition of dye-leakage into conjunctiva 4 hours after injection of carrageenan or LPS, since 59% or 83% of dye-uptake were inhibited, respectively. 1% of indomethacin eye drops showed only a minimal effect. Conclusions. These results suggest that the mechanism for anterior ocular inflammation may be due to up-regulation of COX-2.