Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Neutrophils play a pivotal role in hemorrhagic shock-induced gastric mucosal injury (Smith et al., 1987). It is widely known that neutrophils accumulate in the gastric mucosa after ischemia-reperfusion injury (Vedder et al., 1988; Kim and Hong, 1995) and in gastritis (Blaser, 1987; Kozol et al., 1990). Furthermore, neutrophils have been traced from the microvasculature through the mucosa to the gastric lumen in case of peptic ulceration (Steer, 1985).

Rebamipide (2-(4-chlorobenzoylamino)-3-[2-(1H)-quinolinon-4-yl]propionic acid) has been reported to exert a preventive effect on the drug-induced gastric ulcer formation by inhibiting neutrophil activation (Ogino et al., 1992) and lipid peroxidation (Yoshikawa et al., 1993). In addition, Suzuki et al. (1994) and Yoshida et al. (1996) demonstrated the preventive effect of rebamipide on the gastric mucosal injury induced by human neutrophils activated by Helicobacter pylori. They suggested that the effect of rebamipide was resulted from its inhibitory actions on the neutrophilic oxidative burst. Recently, we reported that rebamipide prevented gastric lesions induced by ischemia-reperfusion via inhibition of the production of reactive oxygen species from activated neutrophils (Kim and Hong, 1995). However, the underlying mechanism(s) by which rebamipide exerts its cytoprotective effect against gastric cell damage is not yet fully determined.

In an in vitro study, we measured the effects of rebamipide on 1) neutrophil-mediated release of BCECF from GMCs, 2) fMLP-stimulated neutrophil aggregation, superoxide anion production and MPO release and 3) binding of [³H]fMLP to neutrophil plasma membranes. In addition, we evaluated in vivo systemic effect of rebamipide on fMLP-induced superoxide anion production and [³H]fMLP binding to its receptors on neutrophils in an ex vivo study.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Gastric mucosal cell culture. Primary cultures of gastric fundic mucosa from newborn New Zealand white rabbits (1-4 wk) were prepared by the method of Terano et al. (1982). In brief, strips of fundic tissue were rinsed three times with BSS and minced into 2 to 3-mm³ pieces with fine scissors in a small amount of BSS. The minced tissues were suspended in BSS containing 0.1% collagenase (Worthington) and 0.05% hyaluronidase. The suspension was incubated at 37°C in a shaking water bath for 60 min. The tissues were pipetted several times to complete dispersion, followed by incubation and pipetting for an additional 15 min, then filtered through a sterile 100-µm nylon mesh (Becton Dickinson Labware, Franklin Lakes, NJ). The filtrate, containing cell clumps, was centrifuged at 600 rpm for 5 min. The sediment was washed twice in BSS by centrifugation at 600 rpm for 5 min. The washed pellet was resuspended in tissue culture medium (Coon's modified Ham's F-12 medium). Confluent monolayers were studied 3 to 4 days after seeding. Histochemical identification of cultured cells was performed using the periodic acid-Schiff reaction for mucus-producing cells and measuring succinate dehydrogenase activity for parietal cells as previously described (Terano et al., 1982).

Isolation of rabbit neutrophils. Neutrophils were isolated from peripheral blood of New Zealand white rabbits (2.5-3.5 kg) with the use of polyvinylpyrrolidone sedimentation and gradient separation on Histopaque 1077. This procedure yielded a neutrophil population that was 95% viable as determined by trypan blue exclusion and 98% pure as determined by acetic acid crystal violet staining.

Cytotoxicity assay. The monolayered GMC were labelled with 5 µM BCECF-AM for 20 min in a 5% CO₂ incubator at 37°C. Cells were then washed three times with 3 ml of culture medium to remove unhydrolyzed BCECF-AM. GMCs were coincubated with neutrophils without (control) or with (vehicle) cytochalasin B (5 µM) plus fMLP (1 nM). The cell mixtures were incubated in 24-well culture plates for 1 hr in a 5% CO₂ incubator at 37°C. For measurement of BCECF (hydrolyzed product from BCECF-AM by intracellular esterases) released from cells, supernatants were harvested and centrifuged to remove floating cells. To determine spontaneous (nonspecific) BCECF release, the following equation was used: percent specific BCECF release = (AB/CB) × 100%; where A represents the mean experimental BCECF release, B represents the mean spontaneous BCECF released and C represents the mean maximum BCECF released. Maximum BCECF release was determined by incubation in 0.05% Triton X-100. Spontaneous BCECF release was determined from the cells after 1 hr of incubation in culture medium.

MPO release. The measurement of fMLP-induced MPO release was performed by incubating 5 × 10⁶ neutrophils in the presence of cytochalasin B (5 µM) at various concentrations of fMLP for 30 min at 37°C. The reaction was terminated by placing the tubes in an ice bath followed by centrifugation.

Superoxide anion production. The amount of superoxide anion produced was determined by the reduction of ferricytochrome C to ferrocytochrome C by stimulated cells as previously described (Babior et al., 1973). Neutrophils (5 × 10⁶) were incubated in the presence of 5 µM cytochalasin B, 60 µM ferricytochrome C with varying concentrations of fMLP for 5 min at 37°C. PMA- and NaF-induced superoxide anion productions were performed in the absence of cytochalasin B, and the cells were incubated at 37°C for 5 and 30 min, respectively. The amount of superoxide anion produced was calculated from the difference in absorbance between samples of cells that received 100 U/ml of superoxide dismutase before activation and those receiving superoxide dismutase after activation.

Neutrophil aggregation. Neutrophils (2 × 10⁶) were resuspended in isotonic N-[2-hydroxyethyl]piperazine-N-[2-ethanesulfonic acid] buffer (Sigma Chemical Co., St. Louis, MO) and warmed to 37°C for 5 min. The change in light transmittance through the stirred (900 rpm) suspensions of cells was monitored in a dual chamber aggregometer (Chrono-Log).

Measurement of [³H]fMLP binding. Binding of [³H]fMLP (New England Nuclear, Boston, MA) to intact neutrophils was performed in triplicate at 37°C in a shaking water bath. Dilutions of [³H]fMLP were made for final concentrations of 40, 20, 10, 7, 5, 3 and 1 nM. Nonspecific binding was defined as the amount of radioligand bound in the presence of 1000-fold excess of unlabelled peptide (10 µM final concentration). Cells (2 × 10⁶) were added to each incubation tube in a final volume of 500 µl. Both cells and radioligand were incubated in 12 × 75 mm polypropylene tubes with and without unlabelled ligand for 30 min at 37°C. Bound label was separated from free label by centrifugation at 12,000 × g for 30 sec. Supernatants were aspirated and the pellet was washed once with ice-cold buffer at 12,000 × g for 30 sec. The cell pellet was solubilized in 0.15 N NaOH and the volume brought to 10 ml with scintillant. Radioactivity was determined using a Packard liquid scintillation counter (TRI-CARB 2100TR).

Drugs. Rebamipide (Otsuka Pharmaceutical Co., Tokushima, Japan) was suspended in 0.5% carboxymethylcellulose and administered orally once a day for 3 days. For the in vitro experiments, rebamipide was dissolved in 10 mM NaOH. Radiolabeled [³H]fMLP (71.5 Ci/mmol) was obtained from New England Nuclear (Boston, MA). BCECF-AM was purchased from Molecular Probes (Eugene, OR). The following chemicals: superoxide dismutase, O-dianisidine hydrochloride, Histopaque 1077, PMA, NaF and fMLP were purchased from Sigma. Other drugs used in this experiment were reagent grade.

Statistics. All data are expressed as means ± S.E.M. Statistical differences between groups were determined by analysis of variance or Student's t test. Differences were significant when P < .05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Assay of cytotoxicity. Injury to GMC was assessed by measuring BCECF release. Spontaneous "nonspecific" BCECF release was normalized, and the results were expressed as "specific BCECF release." Specific BCECF release from nonstimulated neutrophils (control) increased by elevating the number of neutrophils (from 1.5 ± 0.9% at 0.1 × 10⁶ cells to 7.2 ± 0.9% at 5 × 10⁶ cells). However, no further increase in specific BCECF release occurred if the incubation time was increased to 4 hr, indicating that GMC injury occurs almost within 1-hr incubation time (fig. 1). Under pretreatment with 1 nM fMLP (with 5 µM cytochalasin B), BCECF release from 5 × 10⁶ neutrophils (vehicle) significantly increased from 7.2 ± 0.9% (without fMLP) to 16.7 ± 1.4% (P < .01). The increased release of BCECF was reduced by preincubation with rebamipide (100, 300 and 1,000 µM) in a concentration-dependent manner (figs. 2 and 3).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Effect of increasing numbers (0.1-5 × 106 cells) of neutrophils on the release of BCECF from rabbit gastric mucosal cells. Each bar represents mean ± S.E.M. of five experiments. Neutrophils were coincubated with cultured gastric mucosal cells with (filled) or without (open) cytochalasin B (CB, 5 µM) plus fMLP (1 nM).

View larger version (K): [in this window] [in a new window]   Fig. 2.   Protective effect of rebamipide against gastric mucosal cell damage induced by fMLP-activated neutrophils. The cytotoxicity of fMLP-activated neutrophils was visualized by fluorescence microphotographs of rabbit gastric mucosal cells labeled with BCECF-AM (5 µM). Control represents gastric mucosal cells incubated with neutrophils without cytochalasin B (5 µM) and fMLP (1 nM). Vehicle represents gastric mucosal cells incubated with neutrophils with cytochalasin B and fMLP without rebamipide. R300 and R1000, vehicle with rebamipide (R300, 300 µM and R1000, 1000 µM). The results are representative findings of five separate experiments.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Gastric mucosal cell damage induced by fMLP-activated neutrophils and the preventive effect of rebamipide. Pretreatment with rebamipide (R100-R1000, 100-1000 µM) significantly attenuated gastric mucosal cell injury induced by fMLP-activated neutrophils. Each bar represents mean ± S.E.M. of five experiments. * P < .01, vs. control. # P < .05; ## P < .01, vs. vehicle. See others in figure 2.

Effect of rebamipide on aggregation and degranulation. As shown in table 1, rebamipide (100-1000 µM) exerted a concentration-dependent inhibition on fMLP- (1 nM) induced aggregation of neutrophils. However, it did not affect PMA-induced neutrophil aggregation. We further examined the effect of rebamipide on fMLP-induced degranulation of neutrophils. Pretreatment of neutrophils with rebamipide for 5 min at 37°C inhibited MPO release at concentrations between 100 and 1000 µM (Table 1).

Effect of rebamipide on neutrophil superoxide anion production. Neutrophils (5 × 10⁶ cells) produced 4.1 ± 0.8 to 20.8 ± 3.0 nmol of superoxide anion in response to 0.1 to 3 nM fMLP and 1.7 ± 0.7 to 19.6 ± 5.8 nmol of superoxide anion in response to PMA (0.1-3 µg/ml). Preincubation of rabbit neutrophils with rebamipide (100-1000 µM) inhibited fMLP-induced superoxide anion production in a concentration-dependent manner. However, this did not occur with PMA- and NaF-induced superoxide anion production (fig. 4).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of rebamipide (R100-R1000, 100-1000 µM) on superoxide anion production by neutrophils stimulated with cytochalasin B (5 µM) plus fMLP (0.1-3 nM), PMA (0.1-3 µg/ml) or NaF (1-20 mM). Each value represents mean ± S.E.M. of four to five experiments.

Effects of rebamipide on [³H]fMLP binding. Rebamipide (100-1000 µM) significantly inhibited [³H]fMLP binding to formyl peptide receptors on the neutrophil membrane (fig. 5). By Scatchard analysis, the maximum binding of [³H]fMLP to the neutrophil membrane that was pretreated with rebamipide (300 µM) was significantly decreased to 0.44 ± 0.05 pmol/2 × 10⁶ cells compared to 0.57 ± 0.07 pmol/2 × 10⁶ cells in vehicle group (P < .05). However, the K_D of the rebamipide-treated (2.56 nM) and vehicle groups (2.38 nM) were not different (fig. 6).

View larger version (47K): [in this window] [in a new window]   Fig. 5.   Effect of rebamipide (R100-R1000, 100-1000 µM) on [3H]fMLP binding to rabbit neutrophils. Rabbit neutrophils (2 × 106 cells) were preincubated with rebamipide for 10 min at 37°C, and thereafter incubated with [3H]fMLP (20 nM) for 30 min. Percent inhibition of [3H]fMLP binding was determined in the presence of rebamipide 100, 300 and 1,000 µM. Each bar represents the mean ± S.E.M. of four experiments.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Scatchard plots for [3H]fMLP binding to intact rabbit neutrophils. The ratio of bound-to-free fMLP is plotted against bound ligand concentration (pmol). In the rebamipide- (R300, 300 µM) treated group, a similar reciprocal slope (KD) is apparent with a significant decrease in the number of maximum binding sites (Bmax).

Effect of systemic administration of rebamipide. Neutrophils isolated from the rabbits that were orally treated with rebamipide (100 mg/kg for 3 days) were tested for their abilities to release superoxide anion in response to fMLP. In response to 1 nM fMLP, neutrophils obtained from rebamipide-treated animals generated 5.5 ± 1.0 nmol of superoxide anion/5 × 10⁶ cells, whereas those from control animals produced 10.1 ± 0.9 nmol of superoxide anion/5 × 10⁶ cells, suggesting that the neutrophils from the rebamipide-treated rabbits had decreased abilities to release superoxide anion into the extracellular medium in comparison to the neutrophils from nontreated animals (fig. 7). In addition, neutrophils isolated from rebamipide-treated rabbits (100 mg/kg for 3 days) had a smaller number of receptor binding sites for fMLP. The maximum binding of [³H]fMLP to neutrophils isolated from rebamipide-treated rabbits was 0.45 ± 0.10 pmol/2 × 10⁶ cells (P < .05), whereas it was 0.59 ± 0.05 pmol/2 × 10⁶ cells to neutrophils isolated from non-treated animals (fig. 8).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Effect of in vivo treatment of rabbit with rebamipide (100 mg/kg for 3 days, orally) on fMLP-induced release of superoxide anion from neutrophils in ex vivo study. The release of superoxide anion from rabbit neutrophils was determined after application of fMLP (0.1-3 nM) in the presence of cytochalasin B (5 µM). The data represent means ± S.E.M. of four experiments. ** P < .01 vs. vehicle.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Effect of in vivo treatment of rabbit with rebamipide (100 mg/kg for 3 days, orally) on [3H]fMLP binding to rabbit neutrophils ex vivo. Bound-to-free ratio is plotted against bound ligand concentration (pmol). In neutrophils isolated from rebamipide-treated rabbits, similar reciprocal slope (KD) is apparent with a significant decrease in the maximum number of binding sites (Bmax). These results were obtained in three separate experiments.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The major findings of the present study were 1) in the in vitro experiment, rebamipide-inhibited fMLP-induced neutrophil aggregation, MPO release and superoxide anion production in association with a decreased release of BCECF from GMC, 2) rebamipide reduced the maximum binding of [³H]fMLP (B_max) to the neutrophil membrane without a change in the K_D and 3) in the ex vivo study, neutrophils isolated from rabbit treated with rebamipide (100 mg/kg, orally, for 3 days) exhibited significantly lower production of superoxide anion in response to cytochalasin B plus fMLP in association with a reduction in the maximum binding of [³H]fMLP (B_max) to the neutrophils.

Neutrophil-mediated cytotoxicity has been demonstrated toward hepatocytes (Mavier et al., 1988), endothelial cells (Varani et al., 1985) and type II pneumocytes (Suttorp and Simon, 1982). A number of studies have demonstrated that neutrophils are capable of damaging gastric mucosal cells by their abilities to aggregate and to release cell disrupting substances such as oxygen free radicals and proteases (Kvietys et al., 1990; Wallace et al., 1990; Kozol et al., 1994). It is well known that neutrophils, when activated by fMLP (the most potent bacterially derived neutrophil chemotaxis peptide; Becker, 1976), show a high cytotoxicity to a variety of tissues including gastric mucosal cells (Kozol et al., 1994), although unstimulated neutrophils have minor effects.

Our experiments indicate that rebamipide inhibits neutrophil functions, such as neutrophil aggregation, MPO release and superoxide anion production from fMLP-stimulated neutrophils. To better understand where rebamipide inhibits the fMLP signal transduction pathways, we further examined the ability of rebamipide to inhibit superoxide formation by PMA and NaF. PMA and NaF stimulate intermediate steps in the phospholipase C-dependent pathway of fMLP signal transduction (Burkey and Webster, 1993). PMA activates protein kinase C in the receptor-dependent mechanism of signal transduction, which leads to phosphorylation of NADPH-oxidase with production of superoxide anion (Nauseef et al., 1991). NaF stimulates receptor-coupled signal transduction at the level of G-protein resulting in increased generation of superoxide anion in neutrophils (Curnutte et al., 1979). In our experiment, rebamipide did not inhibit superoxide anion production mediated by the postreceptor-dependent agonists, PMA and NaF (Burkey and Webster, 1993; Bokoch, 1995) in contrast to fMLP. Based on these results, the action site of rebamipide is considered to be at the fMLP receptor level specifically.

To further identify the underlying mechanism by which rebamipide exerts its inhibitory effect on fMLP-stimulated neutrophil functions such as aggregation, release of MPO and superoxide anion production, we examined effect of rebamipide on [³H]fMLP binding to the neutrophil membrane. The Scatchard analysis demonstrates that rebamipide consistently decreased binding of [³H]fMLP to its receptors on neutrophils and provides strong evidence to support our hypothesis that the decreased GMC cytotoxicity by rebamipide is ascribed to the decreased production of superoxide anion and MPO release via its inhibitory action on neutrophil fMLP receptors.

Consistent with the in vitro study, neutrophils isolated from rabbits orally treated with rebamipide exhibited decreased production of superoxide anion. Moreover, the suppressed superoxide production was in accordance with the decrease in the number of binding sites for neutrophil fMLP receptors. With these findings it is postulated that rebamipide suppresses chemotactic ligand binding at the neutrophil membrane receptor level and rebamipide is characteristically acting on formyl peptide receptors and it does not scavenge superoxide anion or interfere with the assay for superoxide anion.

In conclusion, it is suggested that the inhibitory effect of rebamipide is related to alterations in the neutrophil membrane that ultimately result in a decrease in the number of binding sites for fMLP to its receptors and thereby inhibits neutrophil functions. Additional studies are required to determine the molecular mechanism(s) by which rebamipide modulates this ligand-receptor interaction.