| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online April 21, 2005 as doi:10.1096/fj.04-3325fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
B
,
,1
* Departments of Medicine and Surgery, Department of Veterans Affairs Medical Center, Long Beach, California, USA;
University of California, Irvine, California, USA; and
Department of Surgery and Science, and Department of Disaster and Emergency Medicine, Graduate School of Medical Science, Kyushu University, Fukuoka, Japan
1 Correspondence: University of California, Irvine DVA Medical Center, 5901 E. Seventh St., Long Beach, CA 90822, USA. E-mail: michael.jones4{at}med.va.gov
SPECIFIC AIMS
Portal hypertension (PHT) is associated with increased susceptibility of the gastric mucosa to injury by a variety of factors including nonsteroidal antiinflammatory drugs (NSAIDs) that nonselectively inhibit both isoforms of cyclooxygenase (COX-1 and -2). PHT gastric mucosa also has excessive nitric oxide (NO) production that contributes to the general increased susceptibility to injury. Although the mechanistic actions of NSAIDs in inhibiting prostaglandin production by modifying the catalytic sites of the COX isozymes or by competing with arachidonic acid substrate for those sites are well characterized, much less is known about the mechanisms that underlie the indirect effects of NSAIDs on COX expression, particularly in the context of various pathophysiological conditions. The aims of our present study were to investigate whether selective COX inhibition is sufficient to cause PHT gastric damage and, if so, to determine the underlying molecular mechanisms involved.
PRINCIPAL FINDINGS
1. Selective inhibition of COX-1 causes damage to PHT, but not normal (normotensive), gastric mucosa
To determine whether selective inhibition of COX-1 and/or COX-2 causes gastric mucosal damage under the condition of PHT, we administered either SC-560, a selective COX-1 inhibitor, or NS-398 or celecoxib, selective COX-2 inhibitors, to rats with experimental PHT or to normal (normotensive) sham-operated (SO) control rats, then assessed the extent of gastric injury compared with injury produced in SO and PHT rats by administration of indomethacin, a nonselective inhibitor of both COX-1 and COX-2. Treatment with indomethacin (40 mg/kg) caused gastric hemorrhagic lesions in both SO and PHT rats with the extent of injury being significantly more severe in PHT vs. SO rats (Fig. 1
A). Treatment with NS-398 (10 mg/kg) or celecoxib (10 mg/kg), a clinically relevant selective COX-2 inhibitor, did not produce gastric mucosal damage in either SO or PHT rats (Fig. 1A
). Although SC-560 treatment did not produce any hemorrhagic lesions in the gastric mucosa of SO rats, treatment with 40 mg/kg SC-560 (a dose that has been shown to retain COX-1 selectivity) did cause significant gastric hemorrhagic lesions in all treated PHT rats that were comparable, within the range of variability, with the lesions induced by 40 mg/kg indomethacin in the SO rats (P=0.06) (Fig. 1A
). The injurious effect of selective COX-1 inhibition on the gastric mucosa of PHT rats was also dose-dependent (Fig. 1B
). Histological assessment of gastric injury reflected these macroscopic findings (histological data are provided in full text article).
|
2. PHT gastric mucosa lacks compensatory COX-2 expression following selective COX-1 inhibition
In normal gastric mucosa, selective inhibition of inducible COX-2 does not cause damage because cytoprotective prostaglandin production is maintained via constitutively expressed COX-1. Selective inhibition of COX-1 also does not cause damage to normal gastric mucosa, presumably as a result of a compensatory increase in COX-2 expression. We examined expression levels of gastric COX-1 and COX-2 in vehicle-treated and SC-560-treated SO and PHT rats. Immunoblot analysis revealed that gastric mucosal COX-2 mRNA and protein expression levels were significantly increased in SC-560-treated SO rats by 2.2-fold (P<0.01) and 1.8-fold (P<0.01), respectively, vs. vehicle-treated SO rats. In contrast, treatment of PHT rats with SC-560 failed to result in increased gastric mucosal COX-2 expression levels relative to vehicle-treated PHT or SO rats. Treatment with SC-560 did not affect gastric COX-1 expression levels in SO or PHT rats (COX expression data provided in full text article).
3. Excessive nitric oxide (NO) production in PHT gastric mucosa underlies the lack of a compensatory increase in COX-2 expression and activity levels in response to selective COX-1 inhibition, by blocking induction of NF-
B, which ultimately results in gastric damage
We previously demonstrated that PHT gastric mucosa has excessive nitric oxide (NO) production, as a result of elevated endothelial NO synthase (eNOS) expression and activity, and that normalization, but not complete inhibition, of NOS activity to the level of normal gastric mucosa using N (
)-Nitro-L-arginine methylester (L-NAME) reduced the increased susceptibility of PHT gastric mucosa to acute injury. As shown in Fig. 2
A, pretreatment in PHT rats for 30 min with L-NAME significantly reduced the gastric damage caused by SC-560 administration in PHT rats by almost 80% (P<0.01). Because excessive NO, acting through inhibition of the NF-B transcription factor, has been shown to inhibit COX-2 expression, we examined whether normalization of NOS activity could act to restore COX-2 expression and activity levels and ameliorate the gastric damage produced by selective COX-1 inhibition in PHT rats. Pretreatment of PHT rats with L-NAME restored the compensatory increase in gastric COX-2 protein expression levels in response to SC-560 (Fig. 2B
). Treatment with L-NAME alone had no effect on gastric mucosal COX-2 protein expression levels (Fig. 2B
). We next investigated the effect of selective COX-1 inhibition on the activation of NF-B in the gastric mucosa of SO and PHT rats. Using an ELISA-based activation assay, we found that the DNA-binding activity of NF-B was significantly increased in nuclear protein extracts from gastric_mucosa of SC-560-treated SO rats vs. vehicle-treated rats by 2.0-fold (P<0.001). In contrast, the NF-B DNA-binding activity obtained from gastric mucosal nuclear extracts of SC-560-treated PHT rats was reduced by 77% (P<0.001) vs. SC-560-treated SO rats. Pretreatment with L-NAME resulted in an induction in NF-B DNA binding activity in gastric mucosal nuclear extracts of SC-560 treated PHT rats that was quantitatively as potent as the induction obtained in gastric mucosal nuclear extracts of SC-560 treated SO rats (P=0.103). We also investigated the effect of the NF-B inhibitor, pyrrolidine dithiocarbamate (PDTC), on COX-2 expression levels and gastric injury in SO rats following administration of SC-560. Pretreatment of SO rats with PDTC completely abolished the increase in NF-B DNA-binding activity following SC-560 administration. Pretreatment of SO rats with PDTC completely abolished the increase in COX-2 expression levels and resulted in significant injury to normotensive gastric mucosa following administration of SC-560. Pretreatment of SO rats with PDTC alone did not affect gastric injury. To verify an effect on COX activity as a result of the differential COX-2 expression levels in the gastric mucosa of PHT vs. SO rats following selective COX-1 inhibition, we measured gastric production of prostaglandin E2 (PGE2) and 6-keto PGF1
(as an indicator of prostacyclin, PGI2) in vehicle-treated and SC-560-treated SO and PHT rats. Administration of SC-560-inhibited gastric PGE2 production in both SO and PHT rats vs. vehicle-treated SO rats by 85% (P<0.005) and 89% (P<0.005), respectively, with no significant difference in extents of inhibition between SC-560-treated SO vs. SC-560-treated PHT rats. Although treatment with SC-560 significantly inhibited gastric 6-keto PGF1 production in both SO and PHT rats, this inhibition was 
30% (P<0.005) less in SO rats, presumably as a result of the compensatory increase in COX-2 levels. Pretreatment of PHT rats with L-NAME prior to administration of SC-560 restored gastric mucosal 6-keto PGF1 production to a rate equivalent to the gastric mucosa of SC-560-treated SO rats.
|
CONCLUSIONS AND SIGNIFICANCE
The mechanisms responsible for the increased susceptibility of portal hypertensive (PHT) gastric mucosa to injury remain unexplored. Our study provides a novel, mechanistic explanation for why selective COX inhibition results in gastric damage which likely also underlies why nonselective NSAIDs cause more severe PHT gastric damage: namely, that excessive nitric oxide production inhibits the induction of NF-B which prevents the compensatory increase in COX-2 expression and activity required to maintain gastric integrity in response to selective COX-1 inhibition (Fig. 3
). Our study also indicates that the use of selective COX inhibitors, particularly the potential use of selective COX-1 inhibitors, should be differentially evaluated for the treatment of PHT vs. normotensive patients. This paradigm likely applies to other pathologic conditions such as arterial hypertension, pulmonary hypertension, and ischemia of the brain and/or kidney where the COX isozymes play important regulatory roles.
|
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-3325fje; doi: 10.1096/fj.04-3325fje
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
