A new mechanism for anti-inflammatory actions of proton pump inhibitors – inhibitory effects on neutrophil–endothelial cell interactions

Authors


Correspondence to: Dr N. Yoshida, First Department of Internal Medicine, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto 602–8566, Japan.
E-mail: nyoshida@koto.kpu-m.ac.jp

Summary

Background: Neutrophil–endothelial cell interactions mediated by adhesion molecules may be involved in gastric mucosal inflammation associated with Helicobacter pylori or nonsteroidal anti-inflammatory drugs.

Aim: To investigate the effects of proton pump inhibitors and histamine-2 receptor antagonists (HRA) on neutrophil-endothelial cell adhesive interactions induced by H. pylori water extract (HPE) or interleukin-1β (IL-1β).

Methods: Human peripheral neutrophils and umbilical vein endothelial cells were incubated with either proton pump inhibitors (lansoprazole and omeprazole) or HRA (famotidine and ranitidine). Neutrophil surface expression of CD11b and CD18 and endothelial cell intercellular adhesion molecule-1 (ICAM-1) and vascular adhesion molecule-1 (VCAM-1) were assessed by flow cytometry and an enzyme immunoassay, respectively. Neutrophil adherence was defined as the ratio of exogenous neutrophils that adhered to the endothelial monolayers.

Results: The expression of CD11b and CD18 on neutrophils and neutrophil-dependent adhesion to endothelial cells elicited by HPE were inhibited by lansoprazole and omeprazole at clinical relevant doses, and the expression of ICAM-1 and VCAM-1 on endothelial cells and endothelial-dependent neutrophil adherence induced by IL-1β were also inhibited by lansoprazole and omeprazole at similar doses. Famotidine and ranitidine had no effect on neutrophil–endothelial cell interactions.

Conclusions: These results indicate that proton pump inhibitors can attenuate neutrophil adherence to endothelial cells via inhibiting the expression of adhesion molecules, suggesting that proton pump inhibitors may have anti-inflammatory activity.

Introduction

Gastric mucosal injury induced by Helicobacter pylori or nonsteroidal anti-inflammatory drugs (NSAIDs) has been attributed to activated polymorphonuclear leukocytes (PMN) that adhere in postcapillary venules and subsequently emigrate into the interstitium. 1–5 PMN–endothelial cell interactions are facilitated by various coordinately regulated adherence complexes that initiate ligand receptor interactions in response to appropriate inflammatory stimuli.6,7 Among the adhesion molecules expressed on leucocytes, the CD11/CD18 integrin family, which is comprised of an immunologically distinct α-subunit (CD11a, CD11b, CD11c) and a common β2-subunit (CD18), plays a major role in the adhesion and transendothelial migration of PMN in vitro and in vivo.7–9 Intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1), which are expressed on activated endothelial cells, play an important role in the adherence of PMN and mononuclear cells, respectively. ICAM-1 is constitutively expressed on endothelial cells and acts as a ligand for CD11/CD18 on PMN, 9 while VCAM-1 acts as a ligand for very late activation antigen-4 (VLA-4) on mononuclear cells. 10 ICAM-1 and VCAM-1 require several hours for maximal surface expression after activation of the endothelial cells.

Lansoprazole (LAN) and Omeprazole (OME) are inhibitors of the H+, K+-adenosine triphosphatase (ATPase) and strongly inhibit the release of H+ from gastric parietal cells. 11 These proton pump inhibitors are more effective in rapid ulcer healing than histamine-2 receptor antagonists (HRA). Recently, it has been suggested that proton pump inhibitors inhibit neutrophil functions such as chemotaxis, superoxide production and degranulation. 12 The objective of the present study was to investigate the effects of two types of acid secretion inhibitors, proton pump inhibitors and HRA, on neutrophil-endothelial cell adhesion elicited by H. pylori extract or interleukin-1β (IL-1β), which is associated with gastric inflammation induced by H. pylori infection.

Materials and methods

H. pylori water extract (HPE)

HPE were prepared as previously described, 3 from the standard strain NCTC 11637. Briefly, the organism was grown on blood agar plates and harvested with sterile cotton swabs into distilled water, using 1.0 mL per plate (109−1010 bacteria). The cell suspension was kept at room temperature for 20 min before centrifugation at 12,000 r.p.m. for 15 min. The resultant supernatant was considered as the initial water extract. No preservatives were added, and the extract was stored at −70°C until assayed. Before use, the water extract was thawed at room temperature and centrifuged at 18,000 r.p.m. for 20 min. To remove much of the high molecular weight material (primarily membrane vesicles and whole flagellae), the supernatant was clarified by passage through a 0.2 micron syringe-adapted filter.

Endothelial cells

Human umbilical vein endothelial cells (HUVEC) were harvested from umbilical cords by collagenase treatment as previously described. 13 The cells were plated in Medium 199 (GIBCO, Grand Island, NY) supplemented with 10% heat-inactivated fetal calf serum (Hyclone Laboratories, Logan, UT), thymidine (2.4 mg/L; Sigma, St. Louis, MO), glutamine (230 mg/L; GIBCO), heparin sodium (10 IU/mL; Sigma), antibiotics (100 IU/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin B; GIBCO) and endothelial cell growth factor (80 μg/mL; Biomedical Technologies, Stoughten, MA). The cultures were incubated at 37°C in a humidified atmosphere with 5% CO2 and expanded by brief trypsinization (0.25% trypsin in phosphate-buffered saline containing 0.02% ethylenediamine tetra-acetic acid). HUVEC of the primary through the third passage were seeded on to 96-well tissue culture plates (GIBCO) coated with 0.1% gelatin and 25 μg/mL fibronectin and used in experiments when confluent.

PMN preparation

Heparinized whole blood was obtained by venipuncture from healthy adult men. The blood was mixed with 6% dextran phosphate-buffered saline (PBS) and incubated for 40 min at room temperature to separate the leucocytes from the red blood cells. PMN were isolated by density gradient centrifugation (400 g for 30 min at 4°C) of the leucocyte-rich plasma using Ficoll-Paque (d = 1.077 μg/mL; Pharmacia, Uppsala, Sweden). After centrifugation, the pellet was suspended in lysis buffer (155 mmol/L ammonium chloride, 10 mmol/L potassium bicarbonate, 0.14 mmol/L ethylenediamine tetra-acetic acid) for 3 min to remove residual red blood cells. The cell suspension was centrifuged (350 g for 3 min at 4°C) and rinsed with Hanks' balanced salt solution (HBSS). This procedure yielded a PMN population that was 95%-98% viable (trypan blue exclusion) and 98% pure (acetic acid-crystal violet staining). The isolated neutrophils were re-suspended in plasma-free HBSS.

Reagents

Monoclonal antibodies (mAbs) to CD11a (G25.2), CD11b (D12), CD18 (L130), ICAM-1 (LB-2) and VCAM-1 (E1/6) were purchased from Becton Dickinson (San Jose, CA). A nonbinding protein (mouse immunoglobulin G, Becton Dickinson) was used as the control in the adhesion assays. LAN, OME, Famotidine (FAM) and ranitidine (RAN) were generously provided by Takeda (Osaka, Japan), Astra Hassle AB (Molndal, Sweden), Yamanouchi (Tokyo, Japan) and Glaxo Wellcome (Middlesex, UK), respectively. These agents were dissolved in ethanol and used at concentrations of 10−6, 10−5 or 10−4 mol/L and contained a final concentration of 0.1% ethanol.

Cell viability

To investigate the effect of proton pump inhibitors or HRA on the viability of PMN or HUVEC, the cells were incubated with LAN, OME, FAM or RAN at a final concentration of 10−4 mol/L for 30 min and 5 h, respectively. Cell viability was assessed by trypan blue dye exclusion.

Flow cytometry

The surface expression of CD11b or CD18 on PMN was determined by using immunofluorescence flow cytometry as previously described. 14 Briefly, the reaction mixture consisted of 106 PMN, HPE (final concentration, 2%) and mAb to CD11b or CD18 (final concentration, 2 μg/mL) in a total volume of 500 μL HBSS. After a 30-min incubation at 37°C, PMN were washed with PBS and fluorescein isothiocyanate (FITC)-conjugated goat F(ab′)2 fragment to mouse immunoglobulin G (Cappel, Durham, NC) was added at a final concentration of 10 μg/mL. After a 15-min incubation at 4°C, the PMN were washed with PBS and analysed on an EPICS Profile flow cytometer (Coulter, Hialeah, FL). To determine the effect of proton pump inhibitors or HRA on the PMN surface expression of CD11b or CD18, these agents were added to a PMN suspension at a final concentration of 10−6, 10−5 or 10−4 mol/L and incubated for 20 min before the addition of the mAb and HPE.

Enzyme immunoassay (EIA)

An EIA was used to assess the binding of mAbs to endothelial cell monolayers as previously described. 15 Confluent HUVEC monolayers were prepared in 96-well plates and incubated for 5 h with IL-1β (20 U/mL) or a combination of IL-1β and proton pump inhibitors or HRA at 10−6, 10−5 or 10−4 mol/L. The cells were fixed by the addition of 1% paraformaldehyde in PBS for 15 min at room temperature, washed three times with PBS and incubated with 2% bovine serum albumin (BSA) for 30 min. After removing the BSA, mAbs directed against ICAM-1 or VCAM-1 were added and incubated at 37°C for 1 h, washed three times in PBS and then incubated for 1 h in peroxidase-conjugated goat affinity-purified F(ab′)2 fragment to mouse immunoglobulin G (Cappel). After washing, substrate (o-phenylenediamine dihydrochloride, 0.4 mg/mL, pH 5.0) was added and incubated for 30 min at room temperature. The plates were then read at 492 nm in a Micro plate reader (Tosoh, Tokyo, Japan) to quantify the amount of bound antibody.

PMN adhesion assays

PMN-dependent adhesion. PMN adhesion to confluent HUVEC monolayer endothelial cells grown in 96-well culture plates was measured by an EIA. 16–17 Briefly, PMN were preincubated with either proton pump inhibitors or HRA (10−6, 10−5 or 10−4 mol/L) for 20 min at room temperature. The PMN suspensions (2 × 105 PMN) containing the acid secretion inhibitors were added to each well along with HPE (final concentration, 2%). After a 30-min incubation at 37°C in a CO2 incubator, the cells were washed three times to remove nonadherent PMN and reagents, and the adherent PMN were fixed with 1% paraformaldehyde in PBS for 15 min and washed. Anti-CD11a mAb was added to the wells and incubated for 30 min, washed and then incubated for 15 min with biotinylated antimouse immunoglobulin in PBS containing carrier protein and 15 mmol/L sodium azide (DAKO, Carpinteria, CA). The wells were washed, horseradish peroxidase-conjugated streptavidin was added for 15 min, and the wells washed again. Subsequently, 200 μL of 0.4 mg/mL ortho-phenylendiamine dihydrochloride (Sigma) in citrate buffer (pH 5) with 0.012% hydrogen peroxide was added. The reaction was stopped by the addition of 50 μL of 1.5 mol/L H2SO4. The plates were read on a micro plate reader (Tosoh) at 492 nm. All experiments were done at least in triplicate. Data were expressed as a percentage of the control group where PMNs were incubated with HPE alone.

Endothelial-dependent adhesion

Confluent HUVEC monolayers prepared in 96-well plates were incubated for 5 h with IL-1β (20 U/mL) or a combination of IL-1β and the acid secretion inhibitors (10−6, 10−5 or 10−4 mol/L). The HUVEC were then washed twice with HBSS and PMN (2 × 105 ) were added to each well. After a 30-min incubation at 37°C in a CO2 incubator, the cells were washed three times to remove nonadherent PMN and reagents, and the adherent PMN were assessed by an EIA as described above.

Statistical analysis

All results are presented as the mean ± standard error. Data were analysed using an analysis of variance ( anova) followed by Scheffé's test, and a P-value less than 0.05 was considered statistically significant.

Results

The incubation of PMN or HUVEC with proton pump inhibitors or HRA resulted in a greater than 96% viability of the cells, indicating that the 0.1% ethanol carrier used for drug dissolution did not affect cell viability (data not shown). Treatment of PMN with HPE resulted in the increased surface expression of CD11b and CD18 ( Fig. 1). While pre-treatment of the PMNs with acid secretion inhibitors did not affect baseline expression of either CD11b or CD18 (data not shown), LAN and OME suppressed the increase in CD11b and CD18 expression induced by HPE in a concentration-dependent manner. However, FAM and RAN did not have any effects on these expressions.

Figure 1.

Effect of acid secretion inhibitors on upregulation of CD11b (A) and CD18 (B) on PMN stimulated by HPE. PMN were incubated with LAN, OME, FAM or RAN for 20 min before the addition of HPE and the surface expression of CD11b and CD18 on PMN was measured by flow cytometry. The results are expressed as a percentage of the control (CONT) where PMN were incubated only with HPE. Data represent the mean ± s.e. of three experiments. #P < 0.05 compared with unstimulated group (UNSTIM) without HPE; *P < 0.05 compared with CONT.

Similar results were obtained when measuring PMN adhesion to HUVEC ( Fig. 2). Thirty minutes after the addition of PMN and HPE to unstimulated HUVEC, PMN adhesion was estimated. HPE increased neutrophil–endothelial cell adhesive interactions and LAN and OME reduced the adherence of HPE-stimulated PMN to HUVEC in a dose-dependent manner, but FAM and RAN had no effect on PMN adhesion.

Figure 2.

Effect of acid secretion inhibitors on the adhesion of HPE-stimulated PMN to HUVEC. PMN pretreated with or without acid secretion inhibitors were added with HPE to unstimulated HUVEC in 96-well culture plates. After a 30-min incubation, the number of adherent PMN was assessed by EIA. The results are expressed as a percentage of control group (CONT) where PMN were incubated only with HPE. Data represent the mean ± s.e. of three experiments performed in triplicate. #P < 0.05 compared with UNSTIM without HPE; *P < 0.01 compared with CONT.

HUVEC were also stimulated with IL-1β for 5 h followed by assessment of the expression of endothelial adhesion molecules on these cells ( Fig. 3). Incubation of the HUVEC with IL-1β resulted in an increased expression of ICAM-1 and VCAM-1. Pre-treatment of the HUVEC with the acid secretion inhibitors did not affect the baseline expression of ICAM-1 and VCAM-1 (data not shown); however, LAN (10−5 and 10−4 M) and OME (10−4 M), but not FAM or RAN, inhibited the IL-1β-induced upregulation of ICAM-1 ( Fig. 3A). VCAM-1 expression was also reduced by the addition of LAN (10−6, 10−5 and 10−4 M) and OME (10−5 and 10−4 M) ( Fig. 3B). The number of adherent PMN to IL-1β-stimulated HUVEC was much greater than to untreated HUVEC ( Fig. 4). Likewise, the adherence of PMN to HUVEC pretreated with both IL-1β and LAN or OME at 10−5 and 10−4 M was inhibited when compared with HUVEC pretreated with IL-1β only, but FAM and RAN did not have any significant influence on PMN adhesion to HUVEC induced by IL-1β.

Figure 3.

Effect of acid secretion inhibitors on surface expression of ICAM-1 (A) or VCAM-1 (B) on IL-1β-stimulated HUVEC. HUVEC were incubated with IL-1β and acid secretion inhibitors for 5 h and the expression of ICAM-1 or VCAM-1 was then assessed by EIA. The results are expressed as the percentage of the control group (CONT) treated with IL-1β alone. Data represent the mean ± s.e. of three experiments. #P < 0.05 compared with UNSTIM group without IL-1β*P < 0.05 compared with CONT.

Figure 4.

Effect of acid secretion inhibitors on PMN adhesion to IL-1β-activated HUVEC. HUVEC in 96-well cultured plates were incubated with IL-1β and acid secretion inhibitors for 5 h, and then PMN (2 × 105 ) were added to each well. After a 30-min incubation, the number of adherent PMN was assessed by EIA. The results are expressed as a percentage of control group (CONT) with HUVEC preincubated with IL-1β. Data represent the mean ± s.e. of three experiments. #P < 0.05 compared with unstimulated group (UNSTIM) without IL-1β*P < 0.05 compared with CONT.

Discussion

The administration of LAN or OME to humans, which significantly inhibits the p-type ATPase of gastric parietal cells, results in the complete inhibition of gastric acid secretion. It has been shown that during the intravenous administration of OME, concentrations of 1–2 × 10−5 mol/L are observed for a period of several hours, 18 while the oral administration of LAN or OME resulted in plasma concentrations of 2–3 × 10−6 mol/L. 19–20 Therefore, in the present study, the proton pump inhibitors were used at final concentrations of 10−6−10−4 mol/L, doses that have been used in other in vitro studies. 12–22

We observed that LAN and OME prevented the surface expression of CD11b and CD18 on PMN and PMN adherence to endothelial cells at clinical relevant doses, both of which were induced by an extract of H. pylori. CD11/CD18 glycoprotein complex-dependent neutrophil adherence has been implicated in gastrointestinal mucosal injury induced by H. pylori, NSAIDs or ischaemia/reperfusion.1–5,23 In previous studies, we found that a water extract of H. pylori, obtained from a patient with asymptomatic gastritis, induced the surface expression of CD11b/CD18 glycoproteins on neutrophils following the adherence and transendothelial migration of neutrophils both in vitro and in vivo.3 Recently, Suzuki et al. reported that OME inhibited neutrophil adherence induced by HPE in rat mesenteric venules using an in vivo microcirculation system. 24 The results of the present in vitro study may explain a mechanism of inhibitory action of OME on neutrophil adhesion. The CD11b/CD18 glycoprotein is stored within specialized granules of resting neutrophils and, when activated, is translocated to the cell surface by granule fusion. 25 Neutrophil activation also triggers the functional activation of preexisting cell surface CD11b/CD18, presumably through conformational and/or topological alterations. 26 Although the exact mechanisms are unclear, it is possible that proton pump inhibitors have a direct effect on either signal transduction pathways or translocation and conformational changes of the CD11b/CD18 complex. Recent studies have demonstrated that proton pump inhibitors may directly inhibit the v-type ATPase (e.g. H+, Ca++-ATPase) of neutrophils as well as the p-type enzyme (e.g. H+, K+-ATPase) in parietal cells, and result in attenuation of superoxide production from neutrophils. 21 The H+, K+-ATPase has been identified as the proton pump of the gastric parietal cell, whereas v-type ATPases work to provide limited acidification of intracellular organelles. 27 OME has been shown to cause a dose-dependent increase in the intralysosomal pH of neutrophils, 21 and Styrt et al. demonstrated that alkalinization of lysosomes interfered with the fusion of cytoplasmic vesicles, which subsequently inhibited NADPH oxidase activation. 28 As upregulation of CD11b/CD18 also requires translocation to the cell surface by granule fusion, the inhibitory effect by proton pump inhibitors in the present study may partially depend on the prevention of lysosomal acidification. While the exact molecular mechanisms by which proton pump inhibitors inhibited the expression of CD11b/CD18 on PMN are unclear, the present results indicate that proton pump inhibitors may protect against gastroduodenal inflammation by inhibiting the infiltration of neutrophils into the extravascular space elicited by infection with H. pylori.

The present study also demonstrated that proton pump inhibitors inhibited the expression of the endothelial adhesion molecules, ICAM-1 and VCAM-1, and endothelial-dependent PMN adhesion, both of which were stimulated by IL-1β. IL-1β is cytokine crucial to the expression of endothelial adhesion molecules induced by H. pylori infection. 29–30 ICAM-1 and VCAM-1 have been reported to play a major role in the intense infiltration of PMN and mononuclear leucocytes in gastric inflammation elicited by H. pylori infection. 30 In addition, neutrophil–endothelial cell interactions mediated by ICAM-1 have been implicated in NSAID-induced gastric mucosal injury.2–4,31 The expression of ICAM-1 and VCAM-1 on endothelial cells requires the activation of transcription factor, nuclear factor-κB, and synthesis of mRNA. 32–33 It is possible that proton pump inhibitors have an effect on the signal transduction for protein synthesis of adhesion molecules. It has been reported that endothelial cells, as well as PMN, are also capable of generating a low pH and may contain the v-type H+-ATPase. 34 In addition, acidification of endocytic compartments by the v-type H+-ATPase in endothelial cells plays an important role in receptor-mediated endocytosis. 34 It is probable that inhibition of the v-type H+-ATPase by proton pump inhibitors may be involved in suppression of adhesion molecule expression on endothelial cells, further studies are necessary however. In clinical studies, some papers have shown that proton pump inhibitors reduced neutrophil infiltration to the mucosa through decreasing H. pylori density in the gastric mucosa, especially that in the antrum. 35–36 However, the administration of proton pump inhibitors to the H. pylori-positive patients has been reported to worsen corpus gastritis associated with the increased neutrophil accumulation. 36–37 The precise mechanisms for these discrepancy between our present in vitro study and clinical results have not been unknown. In H. pylori-infected patients, many kinds of inflammatory mediators, such as IL-8, monocyte chemoattractant protein−1 (MCP-1), IL-1, TNF-α or H. pylori-derived chemoattractant, are involved in the infiltration of inflammatory cells. In the present study, only HPE or IL-1 was used as a stimulant of neutrophils or endothelial cells. The effect of proton pump inhibitors on neutrophil-endothelial cell adhesion induced by other mediators needs to be investigated.

In summary, we have shown that proton pump inhibitors can attenuate the surface expression of adhesion molecules on neutrophils and endothelial cells, and subsequent leukocyte adhesion to endothelial cells. Alternatively, HRA had no effects on PMN-endothelial cell adhesive interactions. Proton pump inhibitors and HRA have been widely used for the treatment of gastric mucosal injury via their strong inhibition of acid secretion. Results of the present study indicate a new mechanism for the anti-inflammatory actions of proton pump inhibitors, namely that proton pump inhibitors may protect against gastric mucosal inflammation induced by H. pylori infection and NSAID administration through modulating leukocyte–endothelial cell interactions. Future studies should determine whether these in vitro proton pump inhibitor anti-inflammatory actions are similar in an in vivo setting.

Ancillary