Mechanisms underlying the protective effects of interleukin 1 in experimental nonsteroidal anti-inflammatory drug gastropathy

GASTROENTEROLOGY 1992;102:1176-1185

Mechanisms Underlying the Protective Effects of Interleukin I in Experimental Nonsteroidal Anti-inflammatory Drug Gastropathy JOHN L. WALLACE, CATHERINE M. KEENAN, MERCEDES CUCALA KENNETH G. MUGRIDGE, and LUCA PARENTE Gastrointestinal Research Group, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada; and Department of Pharmacology, Sclavo Research Laboratories, Siena, Italy

Interleukin 1 (IL-1) has been shown to reduce the severity of experimental gastroduodenal ulceration, but the mechanism of action is unclear. The present study examined the possibility that the mechanism underlying the protective effects of IL-1 in experimental nonsteroidal anti-inflammatory drug (NSAID)-induced gastropathy is related to effects on gastric acid secretion, on prostaglandin synthesis, and/or on neutrophil function. IL-la and IL1~ dose-dependently (1-10 ~g/kg) reduced the severity of gastric damage induced by indomethacin, whereas tumor necrosis factor a (1-10 pg/kg) had no effect. These effects of IL-1 were not completely attributable to a reduction in the volume or acidity of gastric secretion during the 1-hour pretreatment period. Whereas IL-la and IL-11) significantly inhibited pentagastrin-stimulated acid secretion, the dose-response relationship and time course of actions suggested that effects on acid secretion did not fully account for the ability of these agents to reduce indomethacin-induced gastric injury. The maximally effective dose of IL-11) (10 Itg/ kg) in terms of reduction of indomethacin-induced gastric injury did not significantly affect gastric prostaglandin synthesis. Neutrophil function was assessed using two in vivo assays. IL-I~ inhibited migration of neutrophils in response to intradermal injections of N-formyl-methionyl-leucyl-phenylalanine and leukotriene B4 (LTB4) and dose-dependently (0.1-10 ltg/kg) inhibited LTB4-induced neutropenia. These effects could be mimicked by dexamethasone (1 mg/kg SC), which inhibited the neutropenic response to LTB4 and significantly (P < 0.001) reduced the severity of indomethacin-induced gastric damage. Both IL-11) and dexamethasone could significantly reduce the extent of histologically detectable leukocyte margination within the gastric mucosal microcirculation after indomethacin administration. The results of this study suggest that effects of IL-1 on gastric acid secretion or prostaglandin synthesis do not fully account for

its ability to reduce the severity of experimental NSAID-induced gastropathy, whereas inhibitory effects of IL-1 on neutrophil function may contribute significantly to its protective actions.

'nterleukin 1 (IL-1) is produced by numerous cell .types, including activated macrophages, monocytes, keratinocytes, mesangial cells, endothelial cells, and epithelial cells, and appears to participate in fever, inflammation, and acute-phase responses) There is emerging evidence that IL-1 modulates the endocrine, nervous, and immune systems. While structurally quite dissimilar, the a and 13 forms of IL-1 have almost identical biological properties (see reference I for review} and appear to act at the same receptor. 2 A recent addition to the plethora of actions of IL-1 is its ability to reduce the severity of gastroduodenal ulceration in a number of animal models2 -6 For example, IL-1]3 was shown to markedly reduce the severity of gastric damage induced in the rat by ethanol, aspirin, or indomethacin 3-6 or the duodenal damage induced in the rat by cysteamine2 The mechanism through which IL-1 reduced the severity of ulceration in these models is not yet clear. It is possible that stimulation of endogenous prostaglandin synthesis by IL-1, as has been shown in many systems, 4'7-1° may contribute to the ability of this cytokine to reduce mucosal injury. Because IL-1 was found to reduce the gastric damage induced by a dose of indomethacin causing near-complete suppression of gastric prostaglandin synthesis, a we suggested that other actions of IL-1 may contribute to its antiulcer effects in that model. However, a contribution of IL-l-stimulated prostaglandin synthesis before the administration of indomethacin could not be ruled out. © 1992 by the American Gastroenterological Association

0016-5085/92/$3.00

April 1992

There are a n u m b e r of other possible mechanisms that might account for the ability of IL-1 to reduce the severity of experimental ulceration. Acid has been suggested to play an important role in m a n y forms of gastric mucosal injury11-13; therefore, it is conceivable that the antisecretory actions of IL-1 could contribute to its protective effects. IL-1 has recently been shown to inhibit gastric acid secretion w h e n administered IP, 3-5 IV, T M or intracisternally, 6 but it is not clear to what extent these actions and the protective actions of IL-1 are interrelated. We have recently reported that experimental nonsteroidal anti-inflammatory drug (NSAID)-induced gastropathy is a neutrophil-dependent process. 15 Rats rendered neutropenic by pretreatment with an antineutrophil serum or methotrexate were found to be significantly more resistant to the gastric damaging effects of indomethacin or naproxen than were normal rats. Furthermore, pretreatment of rabbits with a monoclonal antibody that prevents neutrophil adherence to the vascular endothelium by binding to the CD11/CD18 adhesion molecule complex was found to prevent indomethacin-induced gastric i n j u r y ) 6 Treatment with the monoclonal antibody also significantly reduced the n u m b e r of marginated leukocytes within the gastric microcirculation. 1BBecause IL-1 has been reported to markedly affect circulating neutrophil numbers and the adherence of neutrophils to the vascular endothelium, 1 it is conceivable that the protective effects of IL-1 against NSAID-induced gastropathy are mediated through such actions. In the present study, we investigated the mechanism through which IL-1 reduces the susceptibility of the rat stomach to damage induced by indomethacin. In particular, we assessed the contribution of the following effects to the protective actions of IL-I: (a) antisecretory effects, (b) effects on endogenous prostaglandin synthesis, and (c) effects on neutrophil function. In some of these experiments, the effects of IL-la and IL-I[~ have been compared with those of tumor necrosis factor a (TNF-a), a cytokine that shares m a n y of the biological actions of IL-11 but does not show antisecretory actions in the rat stomach. 4'14 The antisecretory effects of the cytokines were assessed in two separate models, one involving determination of luminal acidity under basal conditions and the other involving determination of inhibitory effects on pentagastrin-stimulated secretion in anesthetized rats. Because we have previously established that experimental NSAID-induced gastropathy can be prevented by an inhibitor of neutrophil adherence, 18 the effects of IL-1 were assessed in two models in which the responses are dependent on neutrophil adherence. Specifically, the effects of IL-1 on the ability of circulating neutrophils to respond to

IL-1 AND EXPERIMENTALNSAID GASTROPATHY 1177

intradermal or IV administration of chemotactic factors were examined.

Materials and Methods Male Wistar rats weighing 200-225 g were used in all experiments. The rats were obtained from Charles River Breeding Farms (Montreal, Quebec, Canada). All experimental procedures were approved by the Animal Care Committee of the University of Calgary.

Effects of IL-1 and TNF on IndomethacinInduced Gastric Damage Gastric damage was induced by oral administration of indomethacin at a dose of 20 mg/kg. The indomethacin was dissolved in 1.25% sodium bicarbonate (20 mg/mL). In previous studies, 15this dose of indomethacin was found to reproducibly elicit macroscopically visible hemorrhagic erosions in the stomach of the normal rat. In all experiments in which indomethacin was administered to elicit mucosal injury, the rats were deprived of food for 18-22 hours and water for 2 hours before the experiment. Rats were killed 3 hours after administration of the indomethacin, and the stomach was removed and opened along the greater curvature. An individual unaware of the treatment scored the extent of macroscopically visible damage by counting and measuring the lengths (in millimeters) of each lesion. The gastric damage score was then calculated as the sum of the lengths of all lesions. Groups of at least five rats each were pretreated (IP) 1 hour before indomethacin administration with either ILla (1-10 pg/kg), IL-I~ (0.1-10 pg/kg), TNF-u (1 or 10 ~g/ kg), or the vehicle [phosphate buffered saline (PBS)]. Additional experiments were performed in which the rats were pretreated with the histamine H 2 receptor antagonist, cimetidine (50 mg/kg IP) i hour before indomethacin administration. To determine the duration of action of IL-I~, an experiment was performed in which groups of five or more rats each received either IL-I~ (10 pg/kg) or the vehicle (PBS) IP at 0.5, 1, 2, 6, 12, or 24 hours before administration of indomethacin (20 mg/kg orally). As in the experiments described above, the rats were killed 3 hours after indomethacin administration for assessment of the extent of damage.

Effects of IL-1 and TNF on Basal Gastric Acid Secretion To determine the effects of IL-1 on basal gastric acid secretion during the 1-hour pretreatment period, five rats which had been deprived of food for the previous 1822 hours and water for the previous 2 hours were given IL-I~) IP at a dose of 4 ~g/kg. Additional groups of five rats each were treated with IL-la (4 ~g/kg), TNF-a (10 ~tg/kg), or cimetidine (50 mg/kg) at the same time. These doses of IL-la, IL-113, and cimetidine were chosen because they produced approximately a 50% reduction in the severity of indomethacin-induced ulceration in the studies described above. One hour later, the rats were killed by an overdose of ether, and a laparotomy was performed. The

1178 WALLACE ET AL.

stomach was clamped at the junctions with the esophagus and duodenum and was then excised. The gastric contents were emptied into a glass tube, the volume was measured gravimetrically, and the amount of titratable acidity was determined by titrating the sample to pH 7.0 with 0.01 mol/L NaOH using an automatic titration system (Metrohm; Brinkmann Instruments, Rexdale, Ontario, Canada).

Effects of IL-1 a n d TNF on PentagastrinStimulated Gastric Acid Secretion To further investigate the ability of IL-1 and TNF to inhibit gastric acid secretion, the effects of the cytokines were assessed using an in vivo perfused stomach preparation in which acid secretion was stimulated with pentagastrin. The rats were deprived of food for 18-22 hours and water for 2 hours before the experiment. Anesthesia was induced by IP administration of urethane [6 m L / k g of a 20% (wt/vol) solution in normal saline], and a tracheostomy was performed. The stomach was continuously perfused with 37°C normal saline (3 mL/h) through an oregastric catheter, and the perfusate was collected from a second polyethylene catheter inserted through the duodenum into the stomach. The esophagus and pylorus were ligated to prevent contamination by saliva or duodenogastric reflux. At the beginning of the experiment, the stomach was flushed with saline to remove any residual matter. At the completion of the surgery, a period of 15 minutes was allowed for stabilization before any perfusates were collected. Thereafter, the perfusate was collected every 30 minutes and titrated to pH 7 with 0.01 mol/L NaOH using an automatic titration system (Metrohm). Results are expressed as microequivalents of acid per 30 minutes. At the end of the stabilization period, the rats received a single bolus injection (IP) of either IL-la (1-10 l~g/kg), ILl 6 (0.1-10 ~g/kg), TNF-a (1 or 10 ~g/kg), or the vehicle (PBS). Thirty minutes later, the rats received a bolus injection of pentagastrin (20 ~tg/kg IV) followed by a continuous infusion (20 lig" kg -~" h-~). The gastric perfusate was collected for four subsequent 30-minute periods. Additional experiments were performed in which a group of five rats received cimetidine (50 m g / k g IP) rather than one of the cytokines. To determine the duration of action of IL-I[~, an experiment was performed in which groups of five rats each received either IL-I~ (10 p.g/kg) or the vehicle (PBS) IP at 0.5, 1, 2, 4, or 12 hours before beginning the gastric perfusion experiment and the stimulation with pentagastrin.

Effects of IL-1 on Gastric Prostaglandin Synthesis The effects of IL-I[~ on gastric synthesis of 6-ketoprostaglandin F~ (PGF~) were examined both before and after administration of indomethacin. Because IL-I~ has been reported to stimulate the synthesis of cyclo-oxygenase, TM it was particularly important to determine if prostaglandin synthesis recovered towards basal levels more quickly in IL-l-pretreated animals than in controls after indomethacin administration. Groups of 13 rats each were

GASTROENTEROLOGYVol. 102, No. 4

treated with IL-I[~ (10 ~tg/kg) or the vehicle (PBS) IP. One hour later, five rats from each group were killed, and a sample of the gastric corpus was excised for determination of 6-keto-PGFl~ synthesis, as described in detail previously. 17 The other eight rats in each group were given indomethacin (20 m g / k g orally) and were killed 3 hours later. Once again, samples of the gastric corpus were excised for determination of 6-keto-PGFl~ synthesis.

Effects of IL-1 a n d TNF on Neutrophil Function: Leukotriene B4-Induced Neutropenia Rats pretreated with indomethacin (5 m g / k g SC) 1 hour earlier were anesthetized with sodium pentobarbital (60 m g / k g IP), and a carotid artery and jugular vein were cannulated. The pretreatment with indomethacin was performed to eliminate any potential contribution of endothelial prostacyclin synthesis to effects observed after IL-1 administration. Blood samples were collected into heparinized capillary tubes (75-p.L volume) at various times over the course of the experiment. These samples were centrifuged for 4 minutes in a hematocrit centrifuge, and the capillary tube was then scored with a file on either side of the buffy coat. The portion of the capillary tube containing the buffy coat was transferred to an Eppendorf tube, frozen on dry ice, then stored at -20°C. Myeloperoxidase (MPO) activity in these samples was subsequently measured using an assay described in detail previously. TM MPO assays were always performed within 1 week of collecting the samples. In each experiment, four basal samples were taken at 5-minute intervals. Leukotriene B4 (LTB4; 1 ~g) was then injected IV through the jugular cannula, and blood samples were drawn from the carotid cannula at 30 seconds and 1, 2, 5, 10, 20, and 30 minutes later. Changes in circulating neutrophil numbers, as measured by MPO activity, were then calculated as a percent of the basal level (mean of the four basal samples). In pilot experiments (n = 5), IV injection of the vehicle for LTB4 (PBS) had no significant effect on circulating neutrophil numbers. To assess the effects of IL-I[~ on LTB4-induced neutropenia, groups of five rats each received an IP injection of this cytokine (0.1-10 ~g/kg IP) or the vehicle I hour before injection of LTB4. In these experiments, additional blood samples were drawn 55 minutes after administration of IL-113or vehicle, so that any effect of the cytokine on circulating neutrophil numbers could be assessed. Additional experiments were performed in which groups of five rats each were treated with IL-I~ (10 Ilg/kg IP) or the vehicle 24 hours before administration of LTB4. Because IL-let and IL-113 had been found to be approximately equipotent as inhibitors of gastric acid secretion, and because of a more limited supply of IL-la, we only examined the effects of IL-I[~ in the experiments involving LTB4-induced neutropenia.

Effects of IL-1 a n d TNF on Neutrophil Function: Migration in Response to Chemotactic Factors Rats were lightly anesthetized with ether. The back was shaved, and, at separate sites, LTB4 (0.1 ~g

April 1992

in 100 I~L),N-formylated-methionyl-leucyl-phenylalanine (FMLP; 10 pg in 100 I~L)and the vehicle (PBS; 100 p.L) were injected intradermally. Two hours later the rats were killed, and the injections sites were excised, weighed, and frozen on dry ice. The samples were stored at -20°C (for no more than I week), and MPO activity was measured, as above. Groups of five to eight rats each were pretreated, IP, with either IL-113(0.1, 1, or 10 ~g/kg), TNF-a (10 ~g/kg), or the vehicle (PBS) 1 hour before performing the intradermal injections. For the reasons outlined above, IL-lcz was not assessed in these experiments.

Effects of Dexamethasone Because the ability of glucocorticoids to interfere with neutrophil adherence to the vascular endothelium is well documented, 19'2° the effects of pretreatment with dexamethasone on susceptibility of the gastric mucosa to indomethacin-induced injury were assessed. Groups of five rats each were treated with dexamethasone (1 mg/kg IP) or the vehicle 2 hours before oral administration of indomethacin (20 mg/kg). Three hours later the rats were killed, and the severity of gastric damage was scored as in the previous studies. To confirm that dexamethasone interfered with the ability of neutrophils to adhere, its effects in the LTB4-induced neutropenia model were assessed. Groups of five rats each were treated with dexamethasone (1 mg/kg IP) or the vehicle 2 hours before the LTB4 challenge.

IL-1 AND EXPERIMENTAL NSAID GASTROPATHY 1179

Materials Recombinant human IL-113was prepared at Sclavo Laboratories. Recombinant human IL-lct and TNF-~ were provided by Dr. Robert Newton of DuPont Pharmaceuticals (Glenolden, PA). LTB4 and the standards and antisera for the 6-keto-PGFl~ radioimmunoassay were obtained from Caymen Chemicals (Ann Arbor, MI). Tritiated 6keto-PGFl~ was obtained from Amersham Canada Ltd. (Oakville, Ontario, Canada). Pentagastrin was purchased from Ayerst Laboratories (Montreal, Quebec, Canada). Dexamethasone was obtained from K-line Pharmaceuticals (Downsview, Ontario, Canada). All other reagents were obtained from Fisher Scientific (Edmonton, Alberta, Canada) or Sigma Chemical Company (St. Louis, MO).

Statistical Analysis All data are expressed as mean _+ SEM. Student's t test, paired t test, one-way analysis of variance (ANOVA), and the Neuman-Keuls Test were used to compare test with control groups. The particular test used in each circumstance is specified in the legends to the figures and tables. For comparison of circulating neutrophil numbers to MPO activity of blood samples, a linear regression analysis was performed. With all statistical analyses, an associated probability (P value) _<5% was considered significant.

Results

Indomethacin-Induced Gastric Damage Histological Studies In an attempt to assess the extent of neutrophil accumulation within the gastric microcirculation following administration of indomethacin, and the effects of IL-I[~ and dexamethasone on this response, the following histological study was performed. Four groups of five rats each were fasted overnight. Two of the groups of rats were pretreated with saline (IP), whereas one group received IL-1[3 (5 I~g/kg IP) and the other dexamethasone (1 mg/kg IP). One of the saline-pretreated groups received saline (orally) 1 hour later. The other three groups received indomethacin (20 mg/kg orally) either I hour (saline and IL-I~) or 2 hours (dexamethasone) later. One hour after administration of indomethacin or saline, the rats were killed, and the stomach was excised. The stomach was pinned out, mucosal side up, on a wax platform and immersed in neutral buffered formalin. The extent of macroscopically visible damage was scored, as described above. Two hours later, two samples (~2 × 8 mm) of the midcorpus were excised and transferred to flesh formalin. These samples were processed by routine techniques and embedded in paraffin. Thick sections (~4 llm) were mounted on glass slides and stained with H&E. The slides were examined under a light microscope (100× objective) and the prevalence of leukocyte margination was scored on a 0-3 scale, with a score of 3 indicative of extensive leukocyte margination, as described previously. TM This scoring was performed on coded slides, to prevent observer bias.

IL-lcz a n d IL-113 d o s e - d e p e n d e n t l y r e d u c e d the s e v e r i t y of gastric d a m a g e i n d u c e d b y i n d o m e t h a c i n (Figure 1). For e x a m p l e , p r e t r e a t m e n t w i t h IL-lcz or IL-I[~ at doses of 10 l l g / k g r e d u c e d the d a m a g e score from 53 _+ 7 seen in controls to 8 _+ 2 and 2 + 1, r e s p e c t i v e l y (both P < 0.01). As s h o w n in Figure 1, IL-I~ a n d IL-I~ w e r e a p p r o x i m a t e l y equieffective in r e d u c i n g the s e v e r i t y of i n d o m e t h a c i n - i n d u c e d damage, a l t h o u g h at a dose of I I~g/kg the r e d u c t i o n of d a m a g e b y IL-113 ( ~ 4 0 % ) was significant w h i l e that b y IL-I~ was not. In contrast, TNF-et did not signific a n t l y affect the susceptibility to i n d o m e t h a c i n - i n d u c e d m u c o s a l i n j u r y at e i t h e r of the doses tested (1 or 10 ~g/kg). C i m e t i d i n e significantly r e d u c e d the s e v e r i t y of i n d o m e t h a c i n - i n d u c e d gastric d a m a g e b y a p p r o x i m a t e l y the same e x t e n t as the 4 ~ g / k g doses of IL-I~ a n d IL-113 (Figure 1).

Basal Gastric Secretion As s h o w n in Table 1, the v o l u m e a n d titratable acidity of gastric juice p r e s e n t 1 h o u r after IP a d m i n i s t r a t i o n of IL-lcz (4 I~g/kg), IL-I~ (4 llg/kg), or TNF-cz (10 I~g/kg) did not differ significantly from that in v e h i c l e - t r e a t e d controls. P r e t r e a t m e n t w i t h c i m e t i d i n e significantly r e d u c e d the titratable acid-

1180

WALLACE ET AL.

GASTROENTEROLOGY Vol. 102, No. 4

**

100

Z 0

0 Oq

W

8o

lad (D

.<

(3 .< Z 0 I-(3

L~ 0 Z 0

,0.

~t LU n,,

"1Z

orK---0

-"10

, 5

o0

1t 80 60

40.

20 / I" "

0 Clm*fldln*

0

(so mw/kg)

5

DOSEOFCYTOKINEGug/kg) ,

Figure 2. Effects of IL-la (© ©), IL-lp (@ e), TNF-a (A A), and cimetidine on pentagastrin-stimulated gastric acid secretion. The cytokines or cimetidine were administered IP 1 hour before a bolus IV injection of pentagastrin (20 pg/kg) followed by a continuous infusion (20 pg. kg -1. h -1) for 2 hours. Results are expressed as the mean ± SEM percent inhibition relative to controls (in vehicle-treated control rats, the mean rate of acid secretion was 117 _+8 pEq/h; n = 10}. Each test group consisted of five to eight rats. Asterisks denote significant differences compared with the control group (*P < 0.05; **P < 0.01; ANOVA and Neuman-Keuls Test).

ity (by ~-'90%), but not the volume, of the gastric juice.

Time Course of Inhibitory Effects on Indomethacin-Induced Damage and Acid Secretion

Stimulated Gastric Secretion In control rats, pentagastrin administration resulted in a m e a n acid secretory rate of 117 + 8 l~Eq/h (n = 10). IL-113 significantly reduced pentagastrin-stimulated acid secretion at doses of 5 or 10 ~Lg/kg (Figure 2), whereas IL-la only produced significant inhibition at the highest dose tested (10 lxg/kg). TNF-a did not significantly affect pentagastrin-stimulated acid secretion. Cimetidine (50 mg/kg) inhibited pentagastrinostimulated acid secretion by approximately 80% (P < 0.01).

The inhibitory effects of IL-I~ (10 llg/kg) on pentagastrin-stimulated acid secretion were apparent for as long as 4 hours but were no longer apparent at 12 hours (Figure 3). Conversely, significant inhibitory effects of IL-lJ3 on indomethacinoinduced gastric

100 ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Table 1. Effects of Cytokines a n d Cimetidine on Basal Gastric Secretion

Vehicle IL-la (4 ~g/kg) IL-113 (4/~g/kg) TNF-a (10/2g/kg) Cimetidine (50 mg/kg)

129.0 147.4 88.8 224.3 57.8

± ± ± ± ±

39.6 62.9 20.5 28.0 31.7

T ,~

Acid

.J 0

~-

60.

N

40-

Z 0

Volume ~L}

(SO ms/ks )

DOSE OF" CYTOKINE (p,g//kg)

Figure 1. Effects of IL-la (© ©), IL-lp (e @), TNF-a (A A), and cimetidine on susceptibility to gastric damage induced by indomethacin (20 m g / k g orally). Each cytokine was administered IP 1 hour before administration of indomethacin. In rats pretreated with the vehicle, the mean damage score was 53 -+ 7 (n = 23). Each point represents the mean ± SEM of at least five experiments. The effects of pretreatment with cimetidine (50 m g / k g IP) are shown by the vertical bar (mean ± SEM; n = 5). Asterisks denote significant differences from the control group (*P < 0.05; **P < 0.01; ANOVA and Neuman-Keuls Test).

Treatment

Cimefldlne

10

20-

~

;

~

Domog.

Titratable acidity ~Eq) 6.51 6.48 3.52 8.90 0.52

± ± ± + +

2.00 2.53 0.44 1.25 0.22 °

NOTE. Gastric juice was collected 1 hour after IP administration of the test substances to conscious, fasted rats. Values represent the mean + SEM of five rats per group. °Significant difference from the vehicle-treated group, P < 0.05; ANOVA and Neuman-Keuls Test.

0

o

~

~

~

a

I'o

i2

HOURSAFTERIL-|p ADMINISTRATION Figure 3. Comparison of the time course of inhibitory effects of IL-lp (10 pg/kg) on pentagastrin-stimulated acid secretion (e) and on indomethacin-induced gastric damage (A). Each point represents the mean ± SEM of at least five rats, expressed as a percentage of the values for the corresponding vehicle-treated control group. Asterisk denotes significant differences from the control group (P < 0.05; ANOVA and Neuman-Keuls Test).

April 1992

IL-1 AND EXPERIMENTAL NSAID GASTROPATHY

injury were still apparent 12 hours after IL-I[~ administration. When the pretreatment time was extended to 24 hours, a protective effect of IL-I~ was no longer observed (control damage score of 87 ± 6 vs. 80 ± 8 in the IL-113 pretreated group; n -- 10 per group).

Prostaglandin Synthesis As shown in Figure 4, pretreatment with IL-I~ at a dose (10 I~g/kg) that produced almost complete protection against indomethacin-induced gastric damage had no significant effect on gastric 6-ketoPGFI~ synthesis, measured I hour later. When prostaglandin synthesis was determined 3 hours after administration of indomethacin (20 m g / k g orally), again, there was no significant difference between the vehicle- and IL-l[3-treated groups.

Neutrophil Function In an initial series of experiments, the validity and reproducibility of our method for estimating circulating neutrophil numbers was compared with direct counting of neutrophils on blood smears. In a series of 17 consecutive experiments, a blood sample was collected for determination of MPO activity, as described above; a second blood sample (10 ~L) was smeared on a glass slide and stained with a modified Wright's stain (LeukoStat; Fisher Scientific), and neutrophil counts were performed under the 100× objective of a light microscope. There was a highly significant correlation between the estimates of

E

[5~

20-

Vehicle IL-I~'

1181

circulating neutrophil numbers obtained with these two techniques (r -- 0.91; P < 0.01). IV administration of LTB4 (1 ~g) to control rats resuited in a transient, but highly significant (P < 0.001), decrease in circulating neutrophil numbers, as measured by MPO activity of buffy coats (Table 2). Pretreatment with IL-I~ (0.1-10 ~g/kg) 1 hour before LTB 4 challenge had no significant effect on basal circulating neutrophil numbers but dose-dependently inhibited the neutropenia induced by LTB4 administration. At doses of I and 10 I~g/kg, IL-113 completely prevented this effect of LTB4. If rats were pretreated with IL-I~ (10 ~g/kg) 24 hours before the LTB4 challenge, a significant (P < 0.05) increase in circulating neutrophil numbers was observed, but the neutropenia induced by administration of LTB4 was of similar magnitude to that seen in vehicle-treated controls. Effects of IL-113 on neutrophil function were also assessed using an in vivo neutrophil migration assay. Intradermal injection of LTB4 (0.1 I~g) or FMLP (10 ~g) resulted, within 2 hours, in marked increases in skin MPO levels (Figure 5). Pretreatment with TNF-~ (10 l~g/kg IP) 1 hour before the intradermal injections had no significant effect on neutrophil migration into the skin in response to the two agonists. However, IL-I~ did significantly inhibit this response. Only the highest dose of IL-I~ tested (10 ~g/ kg) significantly inhibited neutrophil migration in response to LTB 4, whereas all three doses of IL-113 (0.1, 1, and 10 ~g/kg) significantly inhibited the response to FMLP. In all cases in which IL-113produced a significant inhibitory effect, the levels of MPO in the excised injection sites did not differ significantly from those in samples from sites where the vehicle was injected.

o'> c

v~

Effects of Dexamethasone

16. \\.

"r-

~Z

As shown in Figure 6, dexamethasone inhibited LTB4-induced neutropenia and significantly reduced the severity of gastric damage induced by oral indomethacin.

12.

>-

,'q

b_ Q_ o

4'

Histological Study

I

~

0 Before Indomethacin

After Indomethacin

Figure 4. Effects of IL-lp on gastric prostaglandin synthesis before and after administration of indomethacin (20 mg/kg orally). Rats received an IP injection with either IL-lp (10 lig/kg) or the vehicle. One hour later, five rats from each group were killed, and gastric tissue was excised for determination of 6-keto-PGFlo synthesis. The other rats (eight per group) were given indomethacin and were killed 3 hours later. Asterisk denotes a significant difference (P < 0.001) compared with the corresponding beforeindomethacin group.

The results of the histological evaluation of leukocyte margination within the gastric microcirculation are outlined in Table 3. Indomethacin administration resulted in significant macroscopic injury, although markedly less in the rats killed I hour after administration than in the rats in studies described above in which damage was assessed 3 hours after indomethacin administration. Indomethacin administration resulted in extensive microvascular disruption, as described in detail previously in the

1182

WALLACE ET AL.

GASTROENTEROLOGY Vol. 102, No. 4

Table 2. Effects of IL-1 on LTB(Induced Neutropenia Treatment

Pretreatment time (h)

PBS IL-1 D (0.1/~g/kg) IL-113 (1 pg/kg) IL-l[3 (10 pg/kg) PBS IL-I~) (10/lg/kg)

1 1 1 1 24 24

MPO (U) Pre-LTB4 3.14 2.60 2.60 2.43 3.17 5.50

± ± ± _ ± ±

0.49 0.78 0.47 0.89 0.53 0.25 c

MPO (U) Post-LTB4 1.10 1.48 2.55 2.60 1.39 1.79

± ± ± ± _ ±

0.12 0.29 0.55 0.69 0.37 0.20

% Neutropenia 65 ± 35 ± 2± -7 ± 56 ± 67 ±

7b

9" 7 6 6° 8°

NOTE. IL-113 or vehicle was administered IP 1 or 24 hours before IV administration of LTB4 (1 pg). Blood samples were taken 5 minutes before (Pre-LTB,J and 30 seconds after (Post-LTB4) the administration of LTB4. Results are expressed as the m e a n ± SEM of five experiments per group. A negative n u m b e r indicates a n increase in circulating neutrophil n u m b e r s (i.e., MPO activity). °P < 0.05, bp < 0.01; Pre-LTB4 vs. Post-LTB4 (paired t test). cp < 0.05 compared with the corresponding PBS-treated group (Student's t test).

rabbit stomach. 18Briefly, there was extensive vascular congestion in the subepithelial region that was frequently associated with apparent adherence of leukocytes to the vascular wall. This phenomenon was rarely observed in rats which received saline instead of indomethacin or in rats pretreated with either IL-113 (5 pg/kg) or dexamethasone (1 mg/kg) before administration of indomethacin (Table 3). Discussion

The results of the present study confirm that IL-113 can dose-dependently reduce the severity of gastric damage induced by indomethacin and further show that this action is shared by IL-la, but not TNF-a. Significant effects of IL-I~) were observed at doses of >_1 I~g/kg, with near-complete prevention of

macroscopically visible damage at a dose of 10 p g / kg. The protective effect of IL-113 persisted for 12 hours after its administration but was no longer evident 24 hours after its administration. The mechanism responsible for this action of IL-1 does not appear to be solely related to effects on acid secretion and appears to be unrelated to effects on gastric prostaglandin synthesis, but may be mediated through effects on neutrophil function. The similar effects of IL-la and IL-I~ observed in these studies is consistent with the hypothesis that these two molecules act through a common receptor. 2 The studies on basal acid secretion showed that

mm v

m m Control E~) IL-i (o.ip,g/kg) ~-~ IL-1 (1 /.~/kg) E ~ IL-1 (10/,~g/kg)

r,o o o ~) <

[7213 TNF ( 1 0 / . ~ / k g )

mv)

3.0

'<

'~ m..,a

o

2.5

1.5

o

i.o

- 60

<

50.

- 50

40

z '" o. 0 e~

-4o _~ ,., Z

30

30

20

20

10.

10

0

_~E 2.01

~

60.

Vehicle

LTB 4

FblLP

Figure 5. Effects of p r e t r e a t m e n t w i t h IL-I[~ or TNF-a o n migration of n e u t r o p h i l s into the s k i n in response to i n t r a d e r m a l injections of sterile saline, LTB4 (0.1 pg), or FMLP (10 pg). Groups of five to eight rats each were treated w i t h IL-1D (0.1, 1, or 10 p g / kg), TNF-a (10 pg), or the vehicle I h o u r before i n t r a d e r m a l injections of the chemotactic factors. The rats were killed 2 h o u r s later, a n d n e u t r o p h i l infiltration was determined by measurem e n t of myeloperoxidase activity in the injection sites. Results are s h o w n as m e a n ± SEM. Asterisk denotes significant differences from the corresponding control group (P < 0.05; ANOVA a n d N e u m a n - K e u l s Test).

I

,<-

0 CONTROL

0.0

,", z

DEXAMETHASONE

Figure 6. Effects of p r e t r e a t m e n t w i t h d e x a m e t h a s o n e on susceptibility to i n d o m e t h a c i n - i n d u c e d gastric damage (I) a n d LTB4-induced n e u t r o p e n i a ([]). In the i n d o m e t h a c i n experiments, groups of 11 rats were treated w i t h d e x a m e t h a s o n e (1 mg/kg) or the vehicle (PBS) IP 2 h o u r s before oral a d m i n i s t r a tion of i n d o m e t h a c i n (20 mg/kg). T h r e e h o u r s later the rats were killed, a n d the extent of macroscopically visible damage was determined. In the n e u t r o p e n i a experiments, groups of five rats were treated w i t h either d e x a m e t h a s o n e or vehicle, as above, 2 h o u r s before IV a d m i n i s t r a t i o n of I pg of LTB4. Blood samples were t a k e n before a n d 30 seconds after the injection for determination of circulating n e u t r o p h i l n u m b e r s (see Materials a n d Methods), a n d the results are expressed as the percent change (decrease). The results are s h o w n as m e a n ± SEM. Asterisks denote significant differences c o m p a r e d w i t h the corresponding control group (*P < 0.05; ***P < 0.001; Student's t test).

April 1992

IL-1 AND EXPERIMENTAL NSAID GASTROPATHY

Table 3. Effects of IL-lfl and Dexamethasone an Indomethacin-Indueed Leukocyte Margination Within the Gastric Mucosal Microcirculation

Treatment Saline + vehicle Saline + i n d o m e t h a c i n IL-I~ + i n d o m e t h a c i n Dexamethasone + indomethacin

Macroscopic damage score

Histological score of m u c o s a l leukocyte margination

0.5 _+ 0.5 27.8 _+ 5.8 ° 3.8 + 3.8

0.3 _+ 0.2 2.3 + 0.3 a 0.9 _+ 0.4

3.0 _+ 2.7

0.3 + 0.3

NOTE. I n d o m e t h a c i n was a d m i n i s t e r e d orally at a dose of 20 m g / kg i h o u r after IP a d m i n i s t r a t i o n of saline or IL-I~ (5 ~g/kg) or 2 hours after IP a d m i n i s t r a t i o n of d e x a m e t h a s o n e (1 mg/kg). Damage was scored 1 h o u r after i n d o m e t h a c i n administration. T h e extent of leukocyte m a r g i n a t i o n w i t h i n the m u c o s a l microcirculation was scored on a 0-3 scale, w i t h 3 representing " e x t e n s i v e . " °Significant difference (P < 0.05) vs. all other groups. Data are presented as the m e a n _+ SEM of five rats per group.

doses of IL-I~ and IL-113 producing ~50% reduction of indomethacin-induced damage did not significantly affect the volume or the acidity of gastric juice. In marked contrast, a dose of cimetidine producing a similar reduction of damage reduced basal acid secretion by about 90%. Therefore, antisecretory effects of IL-1 are unlikely to completely account for its ability to reduce the severity of experimental NSAID-gastropathy, although such effects, in concert with other actions of IL-1, could contribute to the protective effects observed in this model. Because effects of IL-1 on basal acid secretion have been shown to be prostaglandin dependent, 4-6 it would seem unlikely that the IL-1 pretreatment would have affected gastric acid secretion after indomethacin administration. The experiments in which the effects of the cytokines on pentagastrin-induced acid secretion were examined further support the hypothesis that antisecretory and protective actions of IL-1 are not interdependent phenomena. As in the case of basal acid secretion, more effective acid suppression could be achieved with cimetidine, but a greater reduction of indomethacin-induced gastric damage could be achieved with IL-1 (u or [~). The time courses for the antisecretory and protective actions of IL-113 also differed markedly. The antisecretory effects of IL-I~ were no longer apparent when the pretreatment time was extended to 12 hours, whereas a significant reduction of indomethacin-induced damage was still evident. The ability of IL-1 to reduce the severity of experimental NSAID-induced gastropathy is not attributable to effects on gastric prostaglandin synthesis. Pretreatment with a dose of IL-113, which reduced gastric damage scores by 97%, had no effect on the

1183

capacity of the gastric corpus to produce 6-ketoPGFI~, the stable hydration production of prostacyclin. IL-I~ has been shown to induce synthesis of cyclo-oxygenasel°; therefore, it is possible that after indomethacin administration, gastric prostaglandin synthesis may return to control levels more rapidly in rats treated with IL-113. However, gastric cyclo-oxygenase activity 3 hours after indomethacin administration, as measured by ability to synthesize 6-ketoPGFI~ in vitro, was not significantly affected by pretreatment with IL-113. The results of the present study are consistent with the hypothesis that NSAID-gastropathy is a neutrophil-dependent process. Adherence of leukocytes to the vascular endothelium occurs soon after contact between NSAIDs and the gastric m u c o s a , 22 apparently contributing to the reduction in gastric mucosal blood flow that follows NSAID administration. 21-23 In the present study, such an effect of indomethacin was apparent by the extensive subepithelial vascular congestion and leukocyte margination, as we have observed previously in a rabbit model of indomethacin-induced gastric injury. 16 In addition, vascular endothelial injury is an early event in the pathogenesis of NSAID-induced gastropathy and appears to be neutrophil-induced, is The critical role of the neutrophil in NSAID-induced gastropathy was shown in two recent studies. Rats rendered neutropenic through treatment with an antineutrophil serum or methotrexate were found to be significantly more resistant to the gastric damaging actions of NSAIDs than were controls. ~5 Furthermore, in a study performed in the rabbit, prevention of neutrophil adherence to the vascular endothelium by treatment with a monoclonal antibody directed against the CD11/CD18 adhesion complex resulted in prevention of indomethacin-induced gastric injury. ~6In the present study, treatment with IL-113significantly reduced the extent of leukocyte margination within the gastric microcirculation following indomethacin administration as well as altering the responses of circulating neutrophils to challenge with chemotactic factors. For example, IL-I[~ dose-dependently reduced the magnitude of LTB4-induced neutropenia. When injected IV, LTB4 causes a rapid increase in the expression of CD11/CD18 on the surface of the neutrophil, leading to adherence to the vascular endothelium and to each other. 24 The net result is a marked, transient decrease in circulating neutrophil numbers. It is noteworthy that complete prevention of LTB4-induced neutropenia was observed when IL113 was administered 1 hour before the LTB4 challenge, but not when administered 24 hours before the LTB~ challenge. The ability of IL-I~ to reduce the severity of indomethacin-induced gastric injury was also no longer apparent when the pretreatment time

1184 WALLACE ET AL.

was extended to 24 hours. The absence of an effect of IL-1 in these models 24 hours after its administration is presumably caused by its inactivation, or perhaps by inactivation of a secondary mediator released in response to IL-1. The experiments involving the effects of IL-1]3 on neutrophil migration into the skin add further support to the hypothesis that the beneficial effects of this cytokine in experimental NSAID-induced gastropathy may be related to effects on neutrophil function. In addition, the current studies showed that the inhibitory effects of IL-I~ on neutrophil responsiveness were not limited to those stimulated by LTB4. IL-113 significantly inhibited the migration of neutrophils in response to LTB4 and to FMLP, although the dose required to produce the effect varied with the agonist used. Only the highest dose of IL-113 tested (10 ~g/kg) significantly inhibited the response to intradermal LTB4, whereas all three doses of IL-113 (0.1, 1, and 10 pg/kg) inhibited the response to intradermal FMLP. Interestingly, TNF-a (10 ~g/kg), which was without significant antiulcer effects, did not significantly affect the response to either agonist in this assay system. If inhibition of neutrophil function contributes to the protective effects of IL-1, then other substances that alter neutrophil function in a similar manner should also reduce the severity of indomethacin-induced gastric injury. The experiments with dexamethasone addressed this hypothesis. Inhibitory effects of glucocorticoids on neutrophil adherence to the vascular endothelium are well established ~9'2e and were confirmed in the present study because dexamethasone pretreatment was found to significantly inhibit the neutropenic response to IV LTB4 administration. Pretreatment with this dose of dexamethasone resulted in a highly significant (P < 0.001) reduction of the extent of indomethacin-induced gastric damage, as well as preventing indomethacininduced leukocyte margination and vascular congestion within the gastric microcirculation. These experiments raise the possibility that IL-1 may act through the release of endogenous glucocorticoids, which could then act to reduce the severity of experimental NSAID-induced gastropathy through a mechanism similar to that of dexamethasone. Indeed, stimulatory effects of IL-1 on the release of corticotropin-releasing factor and adrenocorticotropin have been reported. 2s-27 If the protective effects of IL-1 are mediated via effects on neutrophil function, how might these effects be produced? One possibility is that IL-1 directly alters the responsiveness of neutrophils to endogenous agonists. There is certainly evidence to suggest that IL-1 can directly alter neutrophil functions, such as secretion and phagocytosis. 28-3°Alter-

GASTROENTEROLOGY Vol. 102, No. 4

natively, IL-1 might act at the level of the vascular endothelium, resulting in the release of substances with inhibitory effects on neutrophils. One such substance is prostacyclin3~.32; however, the present results suggest that cyclo-oxygenase products do not mediate the protective actions of IL-1. Another endothelium-derived factor that inhibits neutrophil adherence and is released in response to stimulation by IL-133 is nitric oxide. This substance has been shown to significantly reduce gastric damage induced by ethanol 34 and to inhibit FMLP-induced neutrophil activation in vitro. 35 Furthermore, inhibitors of nitric oxide synthesis have recently been shown to markedly increase leukocyte adherence in the ruesenteric vasculature. 36 The potential role of nitric oxide in the mechanism of action of IL-l-induced protection in experimental NSAID-gastropathy warrants further investigation. In summary, IL-1 exerts potent protective actions in experimental NSAID-induced gastropathy that appear to be unrelated to effects on prostaglandin synthesis and are unlikely to be completely attributable to inhibitory effects on acid secretion. IL-1 was found to significantly depress the responsiveness of circulating neutrophils to stimulation with chemotactic factors and to reduce the extent of histologically detectable leukocyte margination within the gastric microcirculation in response to indomethacin administration. Because inhibition of neutrophil adherence and depletion of neutrophils has previously been shown to reduce the severity of experimental NSAID-induced gastropathy, the results of this study are consistent with the hypothesis that effects of IL-1 on neutrophil function are an important component of the protective mechanism. References 1. Dinarello CA. Cytokines: interleukin-1 and tumour necrosis factor (cachectin). In: Gallin JI, Goldstein IM, Snyderman R, eds. Inflammation: basic principles and clinical correlates. New York: Raven, 1988:195-208. 2. Hannum CH, Wilcox CJ, Arend WP, Joslin FG, Dripps DJ, Heimdal PL, Armes LG, Sommer A, Eisenberg SP, Thompson RC. Interleukin-1 receptor antagonist activity of a human interleukin-1 inhibitor. Nature 1990;343:336-340. 3. Wallace JL, Keenan CM, Mugridge KG, Parente L. Reduction of the severity of experimental gastric and duodenal ulceration by interleukin-l[3. Eur J Pharmacol 1990;186:279-284. 4. Robert A, Olafsson AS, Lancaster C, ZhangW. Interleukin-1 is cytoprotective, antisecretory, stimulates PGE2 synthesis by the stomach, and retards gastric emptying. Life Sci 1991;48:123-134. 5. Uehara A, Okumura T, Sekiya C, Okamura K, Takasugi Y, Namiki M. Interleukin-1 inhibits the secretion of gastric acid in rats: possible involvement of prostaglandin. Biochem Biophys Res Comm 1989;162:1578-1584. 6. Saperas ES, Yang H, Rivier C, Tache Y. Central action of recombinant interleukin-1 to inhibit acid secretion in rats. Gastroenterology 1990;99:1599-1606.

April 1992

IL-1 AND EXPERIMENTAL NSAID GASTROPATHY 1185

7. Breviario R, Proserpio P, Bertocchi F, Lampugnani MG, Mantovani A, Dejana E. Interleukin-1 stimulates prostacyclin production by cultured human endothelial cells by increasing arachidonic acid mobilization and conversion. Arteriosclerosis 1990;10:129-134. 8. Cominelli F, Nest CC, Dinarello CA, Gentilini P, Zipser RD. Regulation of eicosanoid production in rabbit colon by interleukin-1. Gastroenterology 1989;97:1400-1405. 9. Mugridge KG, Donati D, Silvestri S, Parente L. Arachidonic acid lipoxygenation may be involved in interleukin-1 induction of prostaglandin biosynthesis. J Pharmacol Exp Ther 1989;250:714-720. 10. Raz A, Wyche A, Siegel N, Needleman P. Regulation of cycloocygenase synthesis by interleukin-1. J Biol Chem 1988;263:3022-3028. 11. Brodie DA, Chase BJ. Evaluation of gastric acid as a factor in drug-induced gastric hemorrhage in the rat. Gastroenterology 1969;56:206-213. 12. Morris GP, Wallace JL. The role of ethanol and of acid in the production of gastric mucosal erosions in rats. Virchows Arch 1981;38:23-38. 13. Pipkin G, Price CA, Parsons ME. Effect of cimetidine on net ion fluxes across the rat gastric mucosa during mucosal damage after gastric ischemia and after intravenous acetylsalicylic acid. Gastroenterology 1984;87:1283-1291. 14. Wallace JL, Cucala M, Mugridge K, Parente L. Secretagoguespecific effects of interleukin-1 on gastric acid secretion. Am J Physiol 1991;261:G559-G564. 15. Wallace JL, Keenan CM, Granger DN. Gastric ulceration induced by non-steroidal anti-inflammatory drugs is a neutrophil-dependent process. Am J Physiol 1990;259:G462-G467. 16. Wallace JL, McKnight GW, Arfors KE. A monoclonal antibody against the CD18 leukocyte adhesion molecule prevents indomethacin-inducedgastric damage in the rabbit. Gastroenterology 1991;100:878-883. 17. Wallace JL. Increased resistance of the rat gastric mucosa to hemorrhagic damage after exposure to an irritant. Role of the "mucoid cap" and prostaglandin synthesis. Gastroenterology 1988;94:22-32. 18. Wallace JL. Glucocorticoid-induced gastric mucosal damage: inhibition of leukotriene, but not prostaglandin biosynthesis. Prostaglandins 1987;34:311-323. 19. Scheimer RP. The mechanisms of antiinflammatory steroid action in allergic diseases. Ann Rev Pharmacol Toxicol 1985;25:381-412. 20. Bowen DL, Fauci AS. Adrenal corticosteroids. In: Gallin JI, Goldstein IM, Snyderman R, eds. Inflammation: basic principles and clinical correlates. New York: Raven, 1988:877-895. 21. Kitahora T, Guth PH. Effect of aspirin plus hydrochloric acid on the gastric mucosal microcirculation. Gastroenterology 1987;93:810-817. 22. Ashley SW, Sonnenschein LA, Cheung LY. Focal gastric mucosal blood flow at the site of aspirin-induced ulceration. Am J Surg 1985;149:53-59. 23. Gana TJ, Huhlewych R, Koo J. Focal gastric mucosal blood flow in aspirin-induced ulceration. Ann Surg 1987;205:399403. 24. Lindstrom P, Lerner R, Palmblad J, Patarroyo M. Rapid adhe-

sire responses of endothelial cells and of neutrophils induced by leukotriene B4 are mediated by leucocytic adhesion protein DIC8. Scand J Immunol 1990;31:737-744. Berkenbosch F, Van Oers J, Del Rey A, Tilders F, Besedovsky H. Corticotropin-releasing factor-producing neurons in the rat activated by interleukin-1. Science 1987;238:524-526. Rothwell NJ. CRF is involved in the pyrogenic and thermogenic effects of interleukin 1[~ in the rat. Am J Physiel 1989;256:E111-E115. Rivier C, Vale W, Brown M. In the rat, interleukin-la and -[~ stimulate adrenocorticotropin and catecholamine release. Endocrinology 1989;125:3096-3102. Borish L, Rosenbaum R, McDonald B, Rosenwasser LJ. Recombinant interleukin-l[3 interacts with high-affinity receptors to activate neutrophil leukotriene B4 synthesis. Inflammation 1990;14:151-162. Ogle JD, Noel JG, Balasurbramaniam A, Sramkoski RM, Ogle CK, Alexander JW. Comparison of abilities of recombinant interleukin-la and -{3 and noninflammatory IL-113 fragment 163-171 to upregulate C3b receptors (CR1) on human neutrophils and to enhance their phagocytic capacity. Inflammation 1990;14:185-194. Smith RJ, Speziale SC, Bowman BJ. Properties ofinterleukin-1 as a complete secretagogue for human neutrophils. Biochem Biophys Res Commun 1985;130:1233-1240. Kainoh M, Imai R, Umetsu T, Hattori M, Nishio S. Prostacyclin and beraprost sodium as suppressors of activated polymorphonuclear leukocytes. Biochem Pharmacol 1990;39: 477-484. Gryglewski RJ, Szczeklik A, Wandzilak M. The effect of six prostaglandins, prostacyclin and iloprost on generation of superoxide anions by human polymorphonuclear leukocytes stimulated by zymozan or formyl-methionyl-leucyl-phenylalanine. Biochem Pharmacol 1987;36:4209-4212. Beasley D, Schwartz JH, Brenner BM. Interleukin 1-induced prolonged a-arginine-dependentcyclic guanosine monophosphate and nitrite production in rat vascular smooth muscle cells. J Clin Invest 1991;87:602-608. MacNaughton WK, Wallace JL. Endothelium-derived relaxing factor (nitric oxide} has protective actions in the stomach. Life Sci 1989;45:1869-1876. McCall T, Whittle BJR, Boughton-Smith NK, Moncada S. Inhibition of FMLP-induced aggregation of rabbit neutrophils by nitric oxide (abstr). Br J Pharmacol 1988;95:517P. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA 1991:88:4651-4655.

25.

26.

27.

28.

29.

30.

31,

32.

33.

34.

35.

36.

Received May 8, 1991. Accepted August 27, 1991. Address requests for reprints to: John L. Wallace, Ph.D., Department of Medical Physiology, University of Calgary, 3330 Hospital Drive N.W., Calgary, Alberta, T2N 4N1, Canada. Supported by grants from the Medical Research Council of Canada (MRC). Dr. Wallace is an MRC Scientist and a recipient of a Medical Scholar award from the Alberta Heritage Foundation for Medical Research (AHFMR). Dr. Cucala is a recipient of a Fellowship from the AHFMR. The autho~s thank Tai Le for his assistance in performing these studies.