Rofecoxib Produces Intestinal but Not Gastric Damage in the Presence of a Low Dose of Indomethacin in Rats

Nonsteroidal anti-inflammatory drugs (NSAIDs) such as indomethacin damage the gastrointestinal tract as an adverse reaction. Although multiple factors are involved in the pathogenesis of NSAID-induced damage, a deficiency of endogenous prostaglandins (PGs) is most important as a background for such lesions (Whittle, 1981; Weissenborn et al., 1985; Takeuchi et al., 1986, 1991, 2002a; Wallace et al., 1990, 2000; Asako et al., 1992; Whittle et al., 1995; Konaka et al., 1999). PG deficiency caused by NSAIDs is due to an inhibition of both cyclooxygenase (COX)-1 and COX-2. Recent studies, including our own, demonstrated that inhibition of both COX-1 and COX-2 is required for NSAID-induced gastrointestinal damage (Wallace et al., 2000; Tanaka et al., 2001, 2002a,c). Furthermore, we also found up-regulation of COX-2 expression in the gastrointestinal mucosa after administration of NSAIDs as well as the selective COX-1 inhibitor SC-560 (Tanaka et al., 2002a,c). These studies suggested the prerequisite of COX-2 expression and its inhibition in the gastrointestinal ulcerogenic responses to NSAIDs.

On the other hand, indomethacin at a low dose (5 mg/kg) inhibited PG production due to COX-1 inhibition, yet it did not produce damage in the gastrointestinal mucosa (Takeuchi et al., 1986). Indomethacin at lower doses reportedly inhibits COX-1 activity more potently than COX-2 activity (Warner et al., 1999). Thus, it is possible that COX-2 inhibitors damage the gastrointestinal mucosa when administered in combination with a low dose of indomethacin.

In the present study, we examined whether rofecoxib, a selective COX-2 inhibitor, provokes damage in the stomach and/or small intestine when a deficiency of endogenous PGs is concurrently induced by a low dose of indomethacin, in relation to the hypermotility response and COX-2 expression.

Materials and Methods

Male Sprague-Dawley rats (220–260 g; Nippon Charles River, Kanagawa, Japan) were used. The gastric ulcerogenic response was examined in animals fasted for 18 h before the experiment (Tanaka et al., 2002a), whereas the intestinal ulcerogenic response was examined in nonfasting animals (Tanaka et al., 2002c). For the former response, animals were kept in cages with raised mesh bottoms and deprived of food but allowed free access to tap water for 18 h before the experiments. Studies were carried out using four to six rats per group and performed under unanesthetized conditions, unless otherwise specified. All experimental procedures described here were approved by the Experimental Animal Research Committee of Kyoto Pharmaceutical University.

Evaluation of Gastric and Intestinal Ulcerogenic Responses. Various COX inhibitors such as indomethacin (a nonselective COX inhibitor; 3 mg/kg), SC-560 (a selective COX-1 inhibitor; 10 or 30 mg/kg), and rofecoxib (a selective COX-2 inhibitor; 10 or 30 mg/kg), either alone or in combination, were given p.o., and the animals were killed 8 and 24 h later, respectively, to examine the gastric ulcerogenic or intestinal response to the above-mentioned treatments. Upon the combined administration, SC-560 or rofecoxib was given 30 min after the administration of indomethacin. The stomach or small intestine was excised and treated with 2% formalin for fixation of the tissue walls. Then, the stomach or small intestine was opened along the greater curvature or antimesenteric attachment, respectively, and examined for damage under a dissecting microscope with square grids (10×). The area (square millimeters) of macroscopically visible lesions was measured, summed for each tissue, and used as a lesion score. The person measuring the lesions did not know the treatments given to the animals. In some cases, the effects of atropine, ampicillin, and aminoguanidine on the intestinal lesions induced by rofecoxib plus indomethacin were examined. Ampicillin (800 mg/kg) was given p.o. twice 24 h and 30 min before the administration of indomethacin, whereas atropine (3 mg/kg) or aminoguanidine (20 mg/kg) was given s.c. 30 min before.

Determination of Mucosal PGE₂ Content. PGE₂ levels in the gastric and small intestinal mucosa were measured after p.o. administration of indomethacin (3 mg/kg), SC-560 (10 or 30 mg/kg), or rofecoxib (10 or 30 mg/kg). Three, 8, or 24 h later, the animals were killed under deep ether anesthesia, and the gastric or small intestinal tissue was isolated, weighed, and placed in a tube containing 100% ethanol plus 0.1 M indomethacin (Futaki et al., 1994). Then, the tissues were homogenized by Polytron homogenizer (IKA, Tokyo, Japan) and centrifuged at 10,000 rpm for 10 min at 4°C. After the supernatant of each sample had been evaporated with N₂ gas, the residue was resolved in assay buffer and used for determination of PGE₂. The concentration of PGE₂ was measured using a PGE₂ enzyme immunoassay kit (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK).

Analysis of COX-1, COX-2, and iNOS mRNA Expression by Reverse Transcription-PCR. The animals, with or without 18 h fasting, were killed under deep ether anesthesia 6 h after p.o. administration of 3 mg/kg indomethacin, 10 or 30 mg/kg SC-560, and 10 or 30 mg/kg rofecoxib, and the stomachs or small intestines were removed, frozen in liquid nitrogen, and stored at -80°C before use. In some cases, 3 mg/kg atropine was given s.c. 30 min before indomethacin, whereas 800 mg/kg ampicillin was given p.o. twice 24 h and 30 min before administration of indomethacin. The tissue samples were pooled from two or three rats for extraction of total RNA, which was prepared by a single-step acid phenol-chloroform extraction procedure by use of TRIzol (Invitrogen, Carlsbad, CA). Total RNA primed by random hexadeoxy ribonucleotide was reverse-transcribed with the Superscript preamplification system (Invitrogen). The sequences of sense and antisense primers for the rat COX-1 were 5′-AACCGTGTGTGTGACTTGCTGAA-3′ and 5′-AGAAAGAGCCCCTCAGAGCTCAG TG-3′, respectively, giving rise to an 887-bp PCR product (Feng et al., 1993). For the rat COX-2, the sequences of sense and antisense primers were 5′-TGATGACTGCCCA ACTCCCATG-3′ and 5′-AATGTTGAAGGTGTCCGGCAGC-3′, respectively, giving rise to a 702-bp PCR product (Tso et al., 1985). The sequences of sense and antisense primers for the rat iNOS were 5′-CGGTTCACAGTCTTGGTGAAAG-3′ and 5′-CAGGTGTTCCC CAGGTAGGTAG-3′, respectively, giving rise to a 651-bp PCR product (Lyons et al., 1992). For the rat glyceraldehyde-3-phosphate dehydrogenase (G3PDH), a constitutively expressed gene, the sequence was 5′-GAACGGGAAGCTCACTGGCATGGC-3′ for the sense primer and 5′-TGAGGTCCACCACCCTGTTGCTG-3′ for the antisense primer, giving rise to a 310-bp PCR product (Feng et al., 1993; Tso et al., 1985). An aliquot of the reverse transcriptase reaction product served as a template in 35 cycles of PCR with 1 min of denaturation at 94°C, 0.5 min of annealing at 58°C, and 1 min of extension at 72°C on a thermal cycler. A portion of the PCR mixture was electrophoresed in 1.8% agarose gel in TAE buffer (40 mM Tris buffer, 2 mM EDTA, and 20 mM acetic acid, pH 8.1), and the gel was stained with ethidium bromide and photographed.

Determination of Gastric and Intestinal Motility. Gastric and intestinal motility was determined using a miniature balloon according to the previously published method (Takeuchi et al., 1986, 2002a). For determination of gastric motility, the balloon and the support catheter were placed into the stomach through an incision of the forestomach under ether anesthesia. The animals were then kept in Bollman cages, and gastric motility was monitored on a Hitachi recorder (model 056; Hitachi, Mito, Japan) using a pressure transducer (model 151-T; Narco Telecare, Houston, TX) and a polygraph device (model 6M-72; San-ei, Tokyo, Japan) after complete recovery from anesthesia. On the other hand, intestinal motility was determined under urethane anesthesia (1.25 g/kg i.p.). The trachea was cannulated to ensure respiration. A midline incision was then made to expose the small intestine, and a thin, saline-filled balloon, made from silicone rubber and attached to a polyethylene catheter, was introduced into the jejunum via a small incision and tied in place avoiding large blood vessels. The volume in the balloon was adjusted to give an initial resting pressure of 5 mm Hg, which was not sufficient to cause active distension of the intestinal wall, and after allowing the preparation to rest for 30 min, intestinal motility was monitored on a recorder (U-228; Tokai-Irika, Tokyo, Japan) as intraluminal pressure changes, through a pressure transducer and polygraph device (Nihon Kodan, Ibaragi, Japan). Indomethacin (3 mg/kg), 10 or 30 mg/kg SC-560, or 10 or 30 mg/kg rofecoxib was given i.d. after basal gastric or intestinal motility had well stabilized, and the motility was measured for 3 h thereafter. In some cases, 3 mg/kg atropine or 20 mg/kg aminoguanidine was given s.c. 2 h after the administration of indomethacin, whereas 800 mg/kg ampicillin was given p.o. 24 h before. Quantitation of gastric or intestinal motility was performed by measuring the area of motility changes in a recording sheet using NIH Image 1.61 (National Institutes of Health, Bethesda, MD), and the data were expressed as the motility index (arbitrary units).

Determination of Enterobacterial Counts. The enterobacteria were enumerated according to a modified version of the method originally described by Reuter et al. (1997). Six hours after administration of 3 mg/kg indomethacin, the animals were killed under deep ether anesthesia, and their small intestines were removed. After the rinsing of each intestine with sterile saline, the mucosa was scraped with glass slides, weighed, and homogenized in 1 ml of sterile phosphate-buffered saline per 100 mg of wet tissue. Aliquots of the homogenate were placed on blood agar and GAM agar (Nissui, Osaka, Japan). Blood agar plates were incubated at 37°C for 24 h under aerobic conditions, whereas GAM agar plates were incubated for 48 h under standard anaerobic conditions (BBL gas pack pouch anaerobic system; BD Biosciences, Hunt Valley, MD). Plates containing ∼10 to 200 colony-forming units (CFU) were examined for numbers of enterobacteria, and the data were expressed as log CFU/g tissue. In some rats, 800 mg/kg ampicillin was given p.o. twice 24 h and 30 min before the administration of indomethacin, whereas 3 mg/kg atropine was given s.c. 30 min before.

Determination of iNOS Activity. The animals were killed under deep ether anesthesia 24 h after the administration of 3 mg/kg indomethacin and 10 mg/kg rofecoxib, either alone or in combination, and their small intestines were removed. After the rinsing of the intestine with cold saline, the mucosa was scraped with glass slides, weighed, and used for the determination of iNOS activity. The iNOS activity was measured by determining the conversion of radiolabeled l-arginine to citrulline, according to the method described by Brown et al. (1992). The tissue was homogenized in ice-cold buffer (50 mM Tris-HCl, 32 mM sucrose, 1 mM dithiothreitol, 10 μg/ml leupeptin, and 2 μg/ml aprotinin), adjusted to pH 7.4 with NaOH, and centrifuged for 20 min at 10,000 rpm at 4°C. The supernatant was incubated for 60 min at 37°C in a reaction buffer containing [³H]l-arginine at 0.5 μCi/ml. The level of activity of constitutively expressed NOS was determined from the difference in the presence and absence of 1 mM EGTA; the activity of iNOS was evaluated in the presence of 1 mM EGTA. Sample protein content was estimated by a Hitachi spectrophotometer (U-2000; Hitachi, Ibaraki, Japan), and the NOS activity was expressed as picomoles of citrulline per minute per milligram of protein.

Preparation of Drugs. The drugs used were indomethacin, ampicillin (Sigma-Aldrich, St. Louis, MO), SC-560 (Cayman Chemical, Ann Arbor, MI), rofecoxib (synthesized in our laboratory), ampicillin, aminoguanidine, atropine (Nakarai Tesque, Kyoto, Japan), and urethane (Tokyo Kasei, Tokyo, Japan). All COX inhibitors were suspended in a hydroxy propyl cellulose solution (Wako Pure Chemicals, Osaka, Japan) whereas other drugs were dissolved in saline. All drugs were prepared immediately before use and administered p.o., i.d., i.p., or s.c. in a volume of 0.5 ml/100 g body weight. Statistics. Data are presented as the mean ± S.E. of four to six rats per group. Statistical analyses were performed using the two-tailed Dunnett's multiple comparison test, and values of P < 0.05 were considered significant.

Results

Gastric and Intestinal Ulcerogenic Response to Indomethacin Plus Rofecoxib. Neither SC-560 nor rofecoxib at 30 mg/kg alone produced any damage in the stomach (Fig. 1A). Similarly, these agents alone at 10 mg/kg did not damage the small intestine (Fig. 1B). However, when these agents were given together, severe damage was induced in both the stomach and intestine at an incidence of 100%. Likewise, indomethacin at 3 mg/kg did not cause any damage in the stomach or intestine. However, when administered together with rofecoxib, the selective COX-2 inhibitor, it produced damage in the small intestine but not the stomach.

The intestinal lesions caused by 3 mg/kg indomethacin plus 10 mg/kg rofecoxib were significantly attenuated by the antibiotic ampicillin (800 mg/kg) and the anticholinergic drug atropine (3 mg/kg) as well as the selective iNOS inhibitor aminoguanidine (20 mg/kg), at doses that reportedly prevent intestinal lesions induced by a high dose of indomethacin (10 mg/kg) (Takeuchi et al., 2002a; Tanaka et al., 2002c) (Fig. 2).

Effects of Various COX Inhibitors on Mucosal PGE₂ Production. Levels of PGE₂ in the normal rat gastric mucosa and small intestine were 15.3 ± 4.1 and 31.1 ± 6.0 ng/g tissue, respectively (Fig. 3, A and B). Indomethacin at 3 mg/kg significantly decreased the mucosal PGE₂ content in the stomach and small intestine when determined 3 h after the administration. Similarly, SC-560 at the doses used significantly inhibited PGE₂ production in these tissues, whereas rofecoxib did not. This inhibitory effect of indomethacin on PGE₂ production was not affected by coadministration of rofecoxib (not shown).

The PGE₂ content remained reduced in the stomach 8 h after the administration of indomethacin, and this effect was not influenced by additional treatment with rofecoxib. However, when the PGE₂ content was measured in the small intestine 24 h after indomethacin treatment, the PGE₂ content was significantly recovered, reaching approximately 70% of control levels, in a rofecoxib-inhibitable manner. In the animals treated with SC-560, the PGE₂ content was significantly recovered in the stomach 8 h later and in the small intestine 24 h later, and these changes were significantly inhibited by the combined administration of rofecoxib, consistent with our previous studies (not shown) (Tanaka et al., 2002a,c).

Effect of Indomethacin on COX-2 mRNA Expression in the Stomach and Small Intestine. We previously reported that NSAID up-regulates the expression of COX-2 in the gastrointestinal mucosa and that inhibition of this enzyme activity is a key to NSAID-induced damage in these tissues (Tanaka et al., 2002a,c). Since indomethacin at 3 mg/kg decreased PGE₂ contents in both the stomach and small intestine, as effectively as SC-560, we examined whether the expression of COX-2 is also up-regulated in these tissues by indomethacin at the nonulcerogenic dose of 3 mg/kg.

The gene expression of COX-2 was negligible in the normal rat stomach and small intestine, although the expression of G3PDH, the house-keeping gene was clearly observed in these tissues (Fig. 4, A and B). Indomethacin at 3 mg/kg induced COX-2 mRNA expression in the small intestine but not the stomach. The up-regulation of COX-2 expression in the small intestine of the indomethacin-treated rats was attenuated by prior administration of 800 mg/kg ampicillin or 3 mg/kg atropine.

Effects of Indomethacin, SC-560, and Rofecoxib on Gastric and Intestinal Motility. The normal rat stomachs contracted at a frequency of ∼16 to 20/10 min with an amplitude of 18.6 ± 3.2 cm H₂O. The gastric motility was markedly enhanced by intragastric administration of 30 mg/kg SC-560, whereas neither 3 mg/kg indomethacin nor 30 mg/kg rofecoxib had any effect (Fig. 5A). By contrast, the intestinal motility was increased by both 3 mg/kg indomethacin and 10 mg/kg SC-560 but not 10 mg/kg rofecoxib given i.d. (Fig. 5B). The enhanced intestinal motility after the indomethacin treatment was completely inhibited by 3 mg/kg, s.c. atropine (Fig. 6). Neither 800 mg/kg, i.d. ampicillin nor 20 mg/kg, s.c. aminoguanidine had any influence on the enhanced intestinal motility caused by indomethacin.

Effect of Indomethacin on Enterobacterial Numbers in the Intestinal Mucosa. The aerobic and anaerobic bacterial numbers in the normal intestinal mucosa were 6.89 ± 0.19 and 7.03 ± 0.13 log CFU/g tissue, respectively (Table 1). Six hours after the administration of 3 mg/kg, p.o. indomethacin, bacterial numbers in both aerobic and anaerobic conditions increased significantly, the values being 8.35 ± 0.30 and 7.89 ± 0.31 log CFU/g tissue, respectively. The antibiotic ampicillin (800 mg/kg, p.o.) or atropine (3 mg/kg, s.c.) significantly prevented the bacterial invasion after indomethacin treatment, and in the former case the numbers of both aerobic and anaerobic bacteria decreasing even below control levels seen in the normal mucosa.

Effect of Indomethacin on iNOS mRNA Expression in Intestinal Mucosa. It is known that NSAIDs up-regulate iNOS expression in the intestinal mucosa (Tanaka et al., 1999, 2002b; Takeuchi et al., 2003) and that NO derived from iNOS plays a pathogenic role in the development of intestinal lesions induced by NSAIDs (Whittle et al., 1995; Konaka et al., 1999; Takeuchi et al., 2002a; Tanaka et al., 2002b). To examine the involvement of iNOS/NO in the pathogenesis of intestinal lesions induced by a low dose of indomethacin plus rofecoxib, we investigated whether indomethacin at 3 mg/kg causes the expression of iNOS mRNA in the small intestine.

As shown in Fig. 7A, reverse transcription-PCR analysis revealed that iNOS mRNA was expressed in the intestinal mucosa 6 h after administration of 3 mg/kg, p.o. indomethacin, although it was not detected in the normal rat intestine with vehicle treatment. Up-regulation of the iNOS mRNA expression caused by indomethacin was clearly prevented by prior administration of 800 mg/kg, p.o. ampicillin as well as 3 mg/kg, s.c. atropine.

Neither 3 mg/kg, p.o. indomethacin nor 10 mg/kg, p.o. rofecoxib alone had any effect on the iNOS activity in the small intestinal mucosa when examined 24 h after the administration (Fig. 7B). However, the combined administration of indomethacin plus rofecoxib markedly increased the iNOS activity, the values reaching approximately 8 times greater than control levels.

Discussion

The present study demonstrated that rofecoxib provoked damage in the small intestine but not the stomach when administered with a low dose of indomethacin (nonulcerogenic dose). The PG deficiency caused by a low dose of indomethacin induces hypermotility and COX-2 expression in the intestine but not the stomach, resulting in damage when COX-2 is inhibited, again supporting the idea that COX-2 expression is a key to induction of damage in the gastrointestinal mucosa.

First, we confirmed our previous findings (Tanaka et al., 2002b,c) that neither SC-560 nor rofecoxib at 10 mg/kg alone caused intestinal damage in nonfasted rats; yet, these agents given in combination produced gross lesions. Similarly, it was confirmed that the combined administration of SC-560 and rofecoxib at 30 mg/kg produced gastric lesions in fasted rat stomachs, although either agent alone did not (Wallace et al., 2000; Tanaka et al., 2001, 2002a). These data strongly support the contention that inhibition of both COX-1 and COX-2 is required for induction of gastrointestinal damage (Wallace et al., 2000; Tanaka et al., 2002a,c). We found in the present study that indomethacin at 3 mg/kg alone did not cause damage in the stomach or intestine; yet, this agent when administered together with the selective COX-2 inhibitor rofecoxib damaged the small intestine but not the stomach. This dose of indomethacin significantly decreased mucosal PGE₂ content in both the stomach and small intestine when determined 3 h after the treatment. Certainly, SC-560 at the doses used significantly inhibited PG production in these tissues, whereas rofecoxib did not. Thus, the effects of indomethacin at a low dose were very much similar to those of SC-560, the selective COX-1 inhibitor. It was previously reported that SC-560 up-regulates the expression of COX-2 in the stomach and intestine, and the PGs derived from COX-2 may play a compensatory role in maintaining the mucosal integrity under PG deficiency caused by COX-1 inhibition (Tanaka et al., 2002a,c). Since indomethacin even at 3 mg/kg decreased the PGE₂ contents as effectively as SC-560, it is expected that COX-2 expression might also be up-regulated in these tissues by this dose of indomethacin. However, indomethacin at 3 mg/kg induced COX-2 expression in the small intestine but not the stomach. These findings are interesting when considering that indomethacin at this dose produced damage in the small intestine but not the stomach, in the presence of rofecoxib. Indeed, the PGE₂ content was reduced by 3 mg/kg indomethacin in both the stomach and intestine, but it recovered significantly 24 h later only in the small intestine. Certainly, this recovery of PGE₂ was prevented when rofecoxib was administered together with indomethacin. The same has previously been shown in the animals after administration of SC-560 with or without rofecoxib (Tanaka et al., 2002a,c). As shown in our previous study, the reduced PGE₂ content in response to SC-560 was significantly recovered in either the stomach and small intestine, at 8 or 12 h after the administration, respectively (Tanaka et al., 2002a,c).

We previously reported that the up-regulation of COX-2 expression after administration of an ulcerogenic dose of indomethacin (10 mg/kg) is prevented by the antibiotic ampicillin, suggesting a role for enterobacteria in this phenomenon (Tanaka et al., 2005). As expected, the COX-2 expression observed after administration of a low dose of indomethacin was also attenuated by prior administration of ampicillin at the dose that prevented intestinal lesions induced by a low dose of indomethacin plus rofecoxib. These results confirmed a causal relationship between bacterial invasion and COX-2 expression and suggested that this event is critical for development of intestinal lesions induced by a low dose of indomethacin plus rofecoxib. We also examined the enterobacterial counts after administration of 3 mg/kg indomethacin. As expected, the number of bacteria in the mucosa was significantly increased after administration of a low dose of indomethacin. Since intestinal damage induced by a low dose of indomethacin plus rofecoxib was significantly prevented by the antibiotic ampicillin as well as the selective iNOS inhibitor aminoguanidine, it is assumed that the pathogenic mechanism of intestinal damage induced by a low dose of indomethacin plus rofecoxib is very much similar to that of NSAID-induced intestinal damage, involving entrobacteria and iNOS/NO as the causative factor.

The expression of COX-2 in the stomach and intestine after administration of NSAIDs was curtailed by atropine at the dose that inhibited gastrointestinal hypermotility in response to NSAIDs, suggesting a causal relationship between the COX-2 expression and hypermotility (Takeuchi et al., 1986, 2002a,b, 2004a,b; Tanaka et al., 2005). This idea was further supported by the findings of Ohno et al. (2004), who showed that the nonselective nitric-oxide synthase inhibitor N^G-nitro l-arginine methyl ester caused intestinal hypermotility as well as up-regulation of COX-2 expression and resulted in intestinal damage when rofecoxib was coadministered with N^G-nitro l-arginine methyl ester. Abnormal contraction of the intestinal wall results in disruption of the unstirred mucus layer over the epithelium, leading to increased mucosal susceptibility to pathogens and irritants (Anthony et al., 1993). Indeed, atropine potently inhibited intestinal hypermotility as well as bacterial invasion and other inflammatory changes in response to indomethacin, resulting in prevention of intestinal damage (Takeuchi et al., 2002a). However, we found in the present study that indomethacin at 3 mg/kg did not affect gastric motility but produced a marked increase in intestinal motility, in an atropine-inhibitable manner. Certainly, the intestinal lesions induced by indomethacin plus rofecoxib were significantly prevented by pretreatment with atropine, similar to ampicillin and aminoguanidine. These results may provide an explanation as to why a low dose of indomethacin did not up-regulate COX-2 expression in the stomach and produced no damage even in the presence of rofecoxib. The question remains why indomethacin at 3 mg/kg caused hypermotility in the small intestine but not in the stomach, although this agent decreased PGE₂ content in these two tissues, similar to SC-560. The latter agent increased motility in both the stomach and small intestine, suggesting that the hypermotility response is associated with a decrease in PG production due to COX-1 inhibition. At present, the reason why a low dose of indomethacin induced the different motility responses in the stomach and small intestine remains unknown, although it may be related to the mechanism for the hypermotility response induced in these tissues by NSAIDs, including indomethacin. The gastric hypermotility response induced by indomethacin at the ulcerogenic dose (10 mg/kg) occurs via vagal-cholinergic mechanisms, since this event is completely inhibited by bilateral vagotomy or prior administration of atropine (Takeuchi et al., 1986, 1990, 1991). By contrast, the intestinal hypermotility response in small intestine was inhibited by atropine but not vagotomy or hexamethonium (unpublished data). It is assumed that the intestinal hypermotility response is due to a deficiency of PGs in the peripheral tissue, whereas the hypermotility response in the stomach may require PG deficiency not only in the peripheral but also in the central nervous system, involving vagal-cholinergic nerves. Further study is certainly required to clarify the mechanism of gastrointestinal hypermotility responses to NSAIDs.

It should be noted that a low dose of indomethacin did not increase the iNOS activity in the intestinal mucosa, despite causing the up-regulation of iNOS mRNA expression. The combined administration of indomethacin plus rofecoxib, however, markedly increased the iNOS activity. Similar results were previously observed using SC-560 with or without rofecoxib in rat small intestines, indicating that the up-regulation of iNOS is associated with the inhibition of COX-1 (Tanaka et al., 2002b). This is understandable, because iNOS expression is triggered by endotoxin released from enterobacteria (Boughton-Smith et al., 1993) and because bacterial invasion is causally related with intestinal hypermotility due to COX-1 inhibition (Takeuchi et al., 2002a, 2004b). SC-560 alone did not increase iNOS activity, although it caused an up-regulation of iNOS expression through enterobacterial invasion and significantly increased iNOS activity only in the presence of rofecoxib. Both a low dose of indomethacin and SC-560 induced COX-2 expression in addition to the up-regulation of iNOS. Thus, it is assumed that PGs derived from COX-2 may inhibit the iNOS activity, and the additional treatment with rofecoxib unmasks the increase in iNOS activity by inhibiting PG production through COX-2 inhibition. Indeed, Kobayashi et al. (2001) reported that COX-2 down-regulated iNOS expression in rat intestinal epithelial cells.

Given the findings in the present study, we concluded that a low dose of indomethacin produced damage in the small intestine but not in the stomach when administered together with rofecoxib. The pathogenic mechanism of intestinal lesions induced by a low dose of indomethacin plus rofecoxib was very much similar to that caused by the ulcerogenic dose of NSAIDs, involving PG deficiency, enterobacteria, iNOS/NO, and hypermotility. PG deficiency caused by a low dose of indomethacin produces hypermotility and COX-2 expression in the small intestine, resulting in damage when COX-2 is inhibited. In the stomach, however, PG deficiency alone causes neither gastric hypermotility nor COX-2 expression, resulting in no damage even in the presence of rofecoxib. It is assumed that the hypermotility response is a key event in the expression of COX-2 and thereby important in the development of mucosal damage in the gastrointestinal tract.