A complex system of interacting mediators exists in the gastric mucosa to s
trengthen its resistance against injury. In this system prostaglandins play
an important role. Prostaglandin biosynthesis is catalysed by the enzyme c
yclooxygenase (COX), which exists in two isoforms, COX-1 and COX-2. Initial
ly the concept was developed that COX-I functions as housekeeping enzyme, w
hereas COX-2 yields prostaglandins involved in pathophysiological reactions
such as inflammation. In the gastrointestinal tract, the maintenance of mu
cosal integrity was attributed exclusively to COX-I without a contribution
of COX-2 and ulcerogenic effects of non-steroidal anti-inflammatory drugs (
NSAIDs) were believed to be the consequence of inhibition of COX-1. Recent
findings, however, indicate that both COX-1 and COX-2 either alone or in co
ncert contribute to gastric mucosal defence. Thus, in normal rat gastric mu
cosa specific inhibition of COX-1 does not elicit mucosal lesions despite n
ear-maximal suppression of gastric prostaglandin formation. When a selectiv
e COX-2 inhibitor which is not ulcerogenic when given alone is added to the
COX-1 inhibitor, severe gastric damage develops. In contrast to normal gas
tric mucosa. which requires simultaneous inhibition of COX-1 and COX-2 for
breakdown of mucosal resistance, in the acid-challenged rat stomach inhibit
ion of COX-I alone results in dose-dependent injury which is further increa
sed by additional inhibition of COX-2 enzyme activity or prevention of acid
-induced up-regulation of COX-2 expression by dexamethasone. COX-2 inhibito
rs do not damage the normal or acid-challenged gastric mucosa when given al
one. However, when nitric oxide formation is suppressed or afferent nerves
are defunctionalized, specific inhibition of COX-2 induces severe gastric d
amage. Ischemia-reperfusion of the gastric artery is associated with up-reg
ulation of COX-2 but not COX-1 mRNA. COX-2 inhibitors or dexamethasone augm
ent ischemia-reperfusion-induced gastric damage up to four-fold, an effect
abolished by concurrent administration of 16,16-dimethyl-PGE(2).(.) Selecti
ve inhibition of COX-1 is less effective. Furthermore, COX-2 inhibitors ant
agonize the protective effect of a mild irritant or intragastric peptone pe
rfusion in the rat stomach, whereas the protection induced by chronic admin
istration of endotoxin is mediated by COX-1. Finally, an important function
of COX-2 is the acceleration of ulcer healing. COX-2 is up-regulated in ch
ronic gastric ulcers and inhibitors of COX-2 impair the healing of ulcers t
o the same extent as non-selective NSAIDs. Taken together, these observatio
ns show that both COX isoenzymes are essential factors in mucosal defence w
ith specific contributions in various physiological and pathophysiological
situations. (C) 2001 Elsevier Science Ltd. All rights reserved.