ABSTRACT
Title
Role of NSAID-activated gene (NAG-1) in the impairing effects of cyclooxygenase inhibitors on gastric ulcer repair
Authors
R. Colucci1, L. Antonioli1, M. Fornai1, N. Bernardini2, C. Ippolito2, C. Scarpignato 3, C. Blandizzi1.
1Interdepartmental Centre for Research in Clinical Pharmacology and Experimental Therapeutics, 2Department of Human Morphology and Applied Biology, University of Pisa, 3Department of Clinical Sciences, University of Parma, Italy
1Interdepartmental Centre for Research in Clinical Pharmacology and Experimental Therapeutics, 2Department of Human Morphology and Applied Biology, University of Pisa, 3Department of Clinical Sciences, University of Parma, Italy
Abstract
Introduction. Non-steroidal anti-inflammatory drugs (NSAIDs), known to act as inhibitors of cyclooxygenase isoforms (COX-1, COX-2), have the potential of altering the reparation processes of gastric mucosa, thus causing delay in ulcer healing. These detrimental actions have been proposed to depend mainly, although not solely, on the blockade of COX-2. Indeed, the effects of cyclooxygenase inhibitors might also result from interactions with cyclooxygenase-independent molecular targets, among which the NSAID-activated gene (NAG-1) promotes apoptotic cell death. The present study investigates the possible involvement of NAG-1 in the impairment of ulcer repair caused by treatment with cyclooxygenase inhibitors.
Methods. Gastric ulcers were induced in adult male Wistar rats (n=6 per group) by subserosal injection of acetic acid (20%, 40 ml). Four days later, animals were treated once daily with intragastric indomethacin (1 mg/kg, COX-1/COX-2 inhibitor), DFU (5 mg/kg, COX-2 inhibitor) and celecoxib (1 mg/kg, COX-2 inhibitor) for 1, 3 or 7 days. Ulcerated animals receiving drug vehicles were employed as controls. Sham-ulcerated animals served as baseline references. At the end of treatments, gastric ulcerated tissues were dissected and processed to assess: 1) ulcer area (mm2); 2) prostaglandin E2 (PGE2, immunoenzymatic assay); 3) NAG-1, COX-1, COX-2, proliferating cell nuclear antigen (PCNA) and cleaved caspase-3 (RT-PCR, immunohistochemistry, western blot).
Results. At day 1, 3 and 7, control ulcer areas were 59.3±4.3, 57.4±3.8 and 18.7±4 mm2, respectively. Immunohistochemistry analysis revealed an increase in COX-2 expression at the level of epithelial cells, myofibroblasts and vessels in the ulcerated area. At day 3 and 7, but not day 1, ulcer healing was delayed by indomethacin and DFU, but not celecoxib (day 3: indomethacin 67.8±3.2 mm2, p<0.05; DFU 74.4±3.9 mm2, p<0.05; celecoxib 59.2±4.3 mm2,p<0.05). At day 1, in control ulcerated tissues, NAG-1 mRNA and protein expression were induced (2.3 and 2.8 folds, respectively), COX-2, PCNA (marker of cell proliferation) and caspase-3 (marker of apoptotic cell death) protein levels were increased (3, 1.5 and 1.8 folds, respectively), COX-1 expression did not vary, and PGE2 levels were 746±35.8 pg/mg. At this time point, NAG-1 protein induction was further enhanced by indomethacin (4.3 folds) and DFU (3.8 folds), but not celecoxib (2.4 folds). Similar trends were observed for caspase-3 (indomethacin 2.7 folds; DFU 2.3 folds; celecoxib 1.4 folds). PCNA expression was increased further by DFU (4.0 folds) or celecoxib (3.5 folds), but it was reduced by indomethacin (0.9 folds). PGE2 levels were markedly reduced by indomethacin (276±59 pg/mg, p<0.05) and, to a lower extent, also by DFU (428±65 pg/mg, p<0.05) and celecoxib (455±57 pg/mg, p<0.05).
Conclusions. The present findings indicate that different cyclooxygenase inhibitors can exert differential delaying effects on gastric ulcer healing, and suggest that NAG-1 plays a significant role in the detrimental actions of some cyclooxygenase inhibitors on ulcer repair. Indeed, the blunting effects of indomethacin and DFU on ulcer healing can be related to their ability to enhance NAG-1 expression and to promote apoptosis. By contrast, the lack of influence of celecoxib on both ulcer repair and NAG-1 induction, taken together with similar profiles of PGE2 inhibition by celecoxib and DFU, suggests that COX-2 blockade alone does not entirely account for the different inihibitory activities of COX-2 blockers on gastric ulcer healing.
Methods. Gastric ulcers were induced in adult male Wistar rats (n=6 per group) by subserosal injection of acetic acid (20%, 40 ml). Four days later, animals were treated once daily with intragastric indomethacin (1 mg/kg, COX-1/COX-2 inhibitor), DFU (5 mg/kg, COX-2 inhibitor) and celecoxib (1 mg/kg, COX-2 inhibitor) for 1, 3 or 7 days. Ulcerated animals receiving drug vehicles were employed as controls. Sham-ulcerated animals served as baseline references. At the end of treatments, gastric ulcerated tissues were dissected and processed to assess: 1) ulcer area (mm2); 2) prostaglandin E2 (PGE2, immunoenzymatic assay); 3) NAG-1, COX-1, COX-2, proliferating cell nuclear antigen (PCNA) and cleaved caspase-3 (RT-PCR, immunohistochemistry, western blot).
Results. At day 1, 3 and 7, control ulcer areas were 59.3±4.3, 57.4±3.8 and 18.7±4 mm2, respectively. Immunohistochemistry analysis revealed an increase in COX-2 expression at the level of epithelial cells, myofibroblasts and vessels in the ulcerated area. At day 3 and 7, but not day 1, ulcer healing was delayed by indomethacin and DFU, but not celecoxib (day 3: indomethacin 67.8±3.2 mm2, p<0.05; DFU 74.4±3.9 mm2, p<0.05; celecoxib 59.2±4.3 mm2,p<0.05). At day 1, in control ulcerated tissues, NAG-1 mRNA and protein expression were induced (2.3 and 2.8 folds, respectively), COX-2, PCNA (marker of cell proliferation) and caspase-3 (marker of apoptotic cell death) protein levels were increased (3, 1.5 and 1.8 folds, respectively), COX-1 expression did not vary, and PGE2 levels were 746±35.8 pg/mg. At this time point, NAG-1 protein induction was further enhanced by indomethacin (4.3 folds) and DFU (3.8 folds), but not celecoxib (2.4 folds). Similar trends were observed for caspase-3 (indomethacin 2.7 folds; DFU 2.3 folds; celecoxib 1.4 folds). PCNA expression was increased further by DFU (4.0 folds) or celecoxib (3.5 folds), but it was reduced by indomethacin (0.9 folds). PGE2 levels were markedly reduced by indomethacin (276±59 pg/mg, p<0.05) and, to a lower extent, also by DFU (428±65 pg/mg, p<0.05) and celecoxib (455±57 pg/mg, p<0.05).
Conclusions. The present findings indicate that different cyclooxygenase inhibitors can exert differential delaying effects on gastric ulcer healing, and suggest that NAG-1 plays a significant role in the detrimental actions of some cyclooxygenase inhibitors on ulcer repair. Indeed, the blunting effects of indomethacin and DFU on ulcer healing can be related to their ability to enhance NAG-1 expression and to promote apoptosis. By contrast, the lack of influence of celecoxib on both ulcer repair and NAG-1 induction, taken together with similar profiles of PGE2 inhibition by celecoxib and DFU, suggests that COX-2 blockade alone does not entirely account for the different inihibitory activities of COX-2 blockers on gastric ulcer healing.