Background
NSAIDs have huge prescription volumes mostly based on the following two benefits, one is an increase of aged patients necessitating NSAIDs prescription to relieve degenerative change-induced pain and the other is an additional trial for either the prevention of colon polyps or the escape from ischemic cardiovascular diseases [
1,
2]. However, the vast use of NSAIDs is limited by troublesome adverse effects such as the gastric erosion/ulcer, complicated bleeding from ulcers, and more serious complications arising at the small intestine and colon. Since the pharmacological action of NSAIDs is through the inhibition of prostaglandin (PG) synthesis via the suppression of cyclooxygenases (COX) [
3], indiscernible diminution of gastroprotective prostaglandin E
2 (PGE
2) is responsible for these gastrointestinal (GI) adverse effects. Though the invention of selective COX-2 inhibitor, coxib, to guarantee GI safety has been suggested as the solving strategy, this solution also needs to improve. Although NSAIDs are used as potent anti-ulcer drugs, the additional uncovered mechanisms of NSAID toxicity [
4,
5] lead us to develop more potent and safer agents.
In the present time, the best choice for preventing NSAIDs-related GI toxicity is either the combination of NSAIDs and proton pump inhibitors (PPIs) or the choice of coxibs [
6,
7]. Since the healing rate of GI ulcers during continuous use of NSAIDs was greater in PPIs group than histamine type 2 receptor antagonist (H2-RA) group, PPIs have been preferred than H2-RA to cope with the adverse effect of NSAIDs [
8,
9]. Besides fundamental acid suppressive actions of PPIs, several functions have been revealed which are the reduction of pro-apoptotic signaling, acid-independent restoration of proliferating and repairing pathways [
10], a reduction in mucosal oxidative damage, healing promoting action, and endoplasmic reticulum stress relieving mechanism [
11‐
16]. Hahm
et al.[
17‐
19] have also reported that PPIs show the potential activities as anticancer therapeutics based on selective induction of apoptosis, anti-angiogenesis against
Helicobacter pylori-associated carcinogenesis, and direct anti-mutagenic actions during tumorigenesis. However the mechanisms responsible for the protective effects of PPIs in NSAIDs-induced gastric damage remain to be determined.
Heme oxygenase-1 (HO-1), an inducible for the first and rate-limiting enzyme of heme degradation, has been known to protect against the cytotoxicity of oxidative stress and apoptotic cell death as well as inflammatory condition [
20,
21]. Fundamental protective effects of HO-1 against inflammation are mediated via anti-oxidative heme degradation, but also associated with the production of the anti-inflammatory mediators, for which redox dependent transcriptional activator, NF-E2-related factor 2 (Nrf2), and its phosphorylation/activation, and oxidation of Kelch-like ECH-associating protein 1 (Keap1) is mechanistically suggested [
22‐
24]. Since the expression of HO-1 has been induced by anti-oxidative, anti-inflammatory, and ischemic relieving responses, in the current study, we hypothesized that the protective effects of PPIs against NSAIDs-induced gastric damage may be related to HO-1 and consequent angiogenesis beyond innate acid suppression. Altogether, our results demonstrate the novel mechanisms that PPIs induce the expression of HO-1 through activating Nrf2/inactivating Keap1 accompanied with the remuneration of ischemic change and the attenuation of inflammatory mediators, thereby facilitating protection against indomethacin-induced gastric damage.
Methods
Materials and cell cultures
Indomethacin was purchased from Sigma Aldrich (Saint Louis, MO) and pantoprazole was provided from Amore Pacific Pharmaceutical Co. (Seoul, Korea). Antibodies for β-actin, HO-1, α-tubulin, Keap1, Nrf2 and VEGF were all obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Normal rat gastric mucosal RGM-1 cells were provided by Prof. Hirofumi Matsui, MD, PhD (Tsukuba Univ., Japan), were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% (v/v) fetal bovine serum, 100 U/ml penicillin. Human umbilical vascular endothelial cells (HUVECs) were cultured in M199 medium (InnoPharma Screen, Seoul, Korea). Cells were maintained at 37°C in a humidified atmosphere containing 5% CO2. Appropriate amounts of RGM-1 cells or HUVECs were seeded and incubated for 24 h, then they were treated with the indicated dose of pantoprazole or indomethacin and incubated for the indicated times. HUVECs were moved to 1% O2 and 5% CO2 hypoxia chamber and incubated for 0–12 h for the in vitro tube formation assay.
Electron spin resonance (ESR) spectroscopy and ROS generation measurement
Various concentrations of pantoprazole added to a total volume of 200 μl containing 0.05 mM FeSO4, 1 mM H2O2, 1 mM 5,5-dimethylpyrroline-N-oxide (DMPO, Sigma Aldrich, Saint Louis, MO), and 50 mM sodium phosphate at pH 7.4 at room temperature. Reactions were initiated by adding H2O2. After incubation for 1 min, aliquots of the reactions were transferred to a quartz cell and the spectrum of DMPO-OH was examined using an ESR spectrophotometer (JES-TE300, JEOL, Tokyo, Japan) under the following conditions: magnetic field, 338.0 ± 5.0 mT; microwave power, 4.95 mW; frequency, 9.421700 GHz; modulation amplitude, 5 mT; sweep time, 0.5 min; and time constant, 0.03 s. Cellular ROS contents were measured by incubating the control or pantoprazole treated RGM-1 cells with 10 μM H2DCF-DA (Invitrogen Life Technologies, Carlsbad, CA) for 30 min. Fluorescence was measured using a confocal laser microscope (LSM710, Carl Zeiss, Oberkochen, Germany).
Western blot analysis
Treated cells were washed twice with PBS and then lysed in ice-cold cell lysis buffer (Cell Signaling Technology) containing 1 mM phenylmethylsulfonyl fluoride (PMSF, Sigma Aldrich). Proteins in lysates were separated by SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes, which were incubated with primary antibodies, washed, incubated with peroxidase-conjugated secondary antibodies, rewashed, and then visualized using an enhanced chemiluminescence (ECL) system (GE Healthcare, Buckinghamshire, UK).
Electrophoretic mobility gel shift assay (EMSA)
Nuclear and cytoplasmic fractions were extracted using NE-PER Nuclear and cytoplasmic reagents (Pierce, Rockford, IL), according to the manufacturer’s instructions. Antioxidant response element (ARE) oligonucleotide probe, 5′-TTT TCT GCT GAG TCA AGG TCC G-3′, and HIF-1α oligonucleotide probe, 5′-TCT GTA CGT GAC CAC ACT CAC CTC-3′, was labeled with [γ-32P] ATP using T4 polynucleotide kinase (Promega, Madison, WI) and separated from unincorporated [γ-32P] ATP by gel filtration using a nick spin column (GE Healthcare). Before adding the 32P-oligonucleotide (1x105 cpm), 10 μg of nuclear extract was kept on ice for 15 min in gel shift binding buffer. To determine the sequence specificity of the NF-κB DNA interaction, we added an excess of unlabeled oligonucleotides. After 20 min of incubation at room temperature, 2 μl of 0.1% bromophenol blue was added, and samples were electrophoresed through 6% non-denaturing PAGE at 150 V in a cold room. Finally, gels were dried and exposed to X-ray film (Kodak, Rochester, NY).
Immunocytochemistry
Treated cells in chamber slides were fixed by 3.7% formaldehyde for 15 min. After washing, cells were blocked in 5% BSA solution containing 0.1% Triton X-100 in PBS for 1 h at room temperature, and then incubated with primary antibody (1:100) for 12 h at 4°C. Cells were then washed 3 times, incubated with secondary antibody (1:300) for 1 h, and then with 4′-6-diamidino-2-phenylindole (DAPI, 100 ng/ml) for 1 min at room temperature. After washing 3 times, cells were mounted with Prolong Gold antifade reagent (Invitrogen Life Technologies, Carlsbad, CA). Fluorescence was visualized under a confocal laser microscope (LSM710, Carl Zeiss).
RNA isolation and quantitative reverse transcription polymerase chain reaction (qRT-PCR)
After treatment, media was removed by suction and cells were washed with Dulbecco’s phosphate-buffered saline (DPBS) twice. RiboEX (Gene All, Seoul, Korea) was added to plates, which were then incubated for 10 min at 4°C. RiboEX was harvested and placed in a 1.5 ml tube, and chloroform was added and gently mixed. After incubation for 10 min in ice, samples were centrifuged at 10,000 × g for 30 min. Supernatants were extracted and mixed with isopropanol, and mixtures were incubated at 4°C for 1 h. After centrifuging at 13,000
g for 30 min, pellet was washed with 70% (
v/
v) ethanol. After allowing the ethanol to evaporate completely, pellets were dissolved in diethylene pyrocarbonate (DEPC)-treated water (Invitrogen Life Technologies, Carlsbad, CA). cDNA was prepared using reverse transcriptase derived from murine Maloney leukemia virus (Promega, Madison, WI), according to the manufacturer’s instructions. PCR was performed over 30 cycles of: 94°C for 20 sec, 58°C for 30 sec, and 72°C for 45 sec. Oligonucleotide primers for PCR (Table
1) were purchased from Bioneer (Daejeon, Korea). All qRT-PCR experiments were repeated in triplicate and quantification was shown in mean ± SD.
Table 1
The sequence of PCR primers
GAPDH | Forward 5′-GGT GCT GAG TAT GTC GTG GA -3′ |
Reverse 5′-TTC AGC TCT GGG ATG ACC TT-3′ |
HO-1 | Forward 5′-GAG AGC ATG TCC CAG GAT TT-3′ |
Reverse 5′-GGT TCT GCT TGT TTC GCT CT -3′ |
COX-2 | Forward 5′-GAA ATG GCT GCA GAG TTG AA -3′ |
Reverse 5′-TCA TCT AGT CTG GAG TGG GA -3′ |
HIF-1α | Forward 5′-AAC AAA CAG AAT CTG TCC TC-3′ |
Reverse 5′-GGT AAT GGA GAC ATT GCC AG-3′ |
VEGF | Forward 5′-CAA TGA TGA AGC CCT GGA GT-3′ |
Reverse 5′-GAT TTC TTG CGC TTT CGT TT -3′ |
PDGF | Forward 5′-AGG AAG CCA TTC CCG CAG TT-3′ |
Reverse 5′-CTA ACC TCA CCT GGA CCT CT -3′ |
bFGF | Forward 5′-TAT GAA GGA AGA TGG ACG GC-3′ |
Reverse 5′-AAC AGT ATG GCC TTC TGT CC -3′ |
IL-1β | Forward 5′-CAT TGT GGC TGT GGA GAA G-3′ |
Reverse 5′-ATC ATC CCA CGA GTC ACA GA -3′ |
IL-8 | Forward 5′-CAG ACA GTG GCA GGG ATT CA-3′ |
Reverse 5′-TTG GGG ACA CCC TTT AGC AT-3′ |
TNF-α | Forward 5′-TAC TGA ACT TCG GGG TGA TT -3′ |
Reverse 5′-CAG CCT TCT CCC TTG AAG AG-3′ |
ICAM-1 | Forward 5′-TGT GCT TTG AGA ACT GTG GC-3′ |
Reverse 5′-GGT TCT GTC CAA CTT CTC AG -3′ |
VCAM-1 | Forward 5′-GAG ACA AAA CAG AAG TGG AAT-3′ |
Reverse 5′-TAC AAG TGG TCC ACT TAT TTC -3′ |
NOX1 | Forward 5′-GAG AAA TTC TCG GAA CTG CC-3′ |
Reverse 5′-TGT TGG CTT CTA CTG TAG CG -3′ |
In vitroangiogenesis assay
This assay was performed using a commercial kit according to the manufacturer’s instruction (Millipore, Billerica, MA). For hypoxic conditions, ECMatrix and Diluent buffer were mixed to make a solid gel, which was then plated in 96 well microplates. Human umbilical vein endothelial cells (HUVEC, 1.0 × 105/ml) were seeded with control, Indomethacin, indomethacin plus pantoprazole, indomethacin plus pantoprazole plus ZnPPIX incubated at 37°C for 4 h. Tube formation was observed under a light microscope.
Indomethacin-induced gastric damage model
A total of 48 rats were purchased from Charles River (Osaka, Japan). Animals were handled in an accredited animal facility in accordance with Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International) guidelines under the facility named CACU (The Center of Animal Care and Use) of Lee Gil Ya Cancer and Diabetes Institute, Gachon University after Institutional Ethics Review Board approval. Animals were divided into four groups each consisting of 12 rats and were starved for 24 h before the experimentation as follows; all rats were administered with intraperitoneal injection of indomethacin (10 mg/kg) to provoke gastric damages, second group with intraperitoneal injection of 10 mg/kg of pantoprazole, the third group with intraperitoneal injection of 30 mg/kg of pantoprazole, and the fourth group with intraperitoneal injection of 30 mg/kg of pantoprazole and 5 mg/kg of ZnPPIX to inhibit the activity of HO-1. Animals were sacrificed 16 h after each administration. The stomachs of rats were removed and opened along the greater curvature and then washed with ice cold phosphate buffered solutions. The numbers of either erosions or ulcers were determined under the magnified photographs. Homogenates obtained from scratched gastric mucosa were kept into liquid nitrogen tank until the assay.
Statistical analysis
Results are expressed as the mean ± SD. The data were analyzed by one-way ANOVA, and the statistical significance between groups was determined by Duncan’s multiple range test. Statistical significance was accepted at P < 0.05.
Discussion
Before the current investigation, several researchers have published that PPIs could prevent gastric mucosal injury by other mechanisms beyond the action of its specific acid inhibition [
11,
14,
25‐
28], we could add more evidences regarding protective action of PPIs against NSAIDs. Pantoprazole significantly induced the expression of HO-1 through Nrf2 activation relevant to anti-inflammatory, anti-oxidative, and ischemia relieving actions. Novel finding of this study different from other investigations is that pantoprazole shows protective actions to the NSAIDs-induced injury through either the correction of angiogenic handicap or the attenuation of inflammation propensity.
Becker JC
et al.[
16] demonstrated that both omeprazole and lansoprazole protected human gastric epithelial and endothelial cells against oxidative stress similar to our study, but used different kinds of PPIs. Since this effect was abrogated in the presence of the HO-1 inhibitor, ZnPPIX, HO-1 seems to be a right target of PPIs in both endothelial and gastric epithelial cells. In this study, we observed untouched novel finding that HO-1 induced by PPIs decreased NSAIDs-incurred inflammation and angiogenic derangement. Takagi T
et al.[
25] also investigated the role of Nrf2, its phosphorylation/activation, and oxidation of Keap1 in lansoprazole-induced HO-1 up-regulation using same cell line with us, RGM-1 cells. When RGM-1 cells were transfected with HO-1 enhancer luciferase reporter plasmid containing mutant stress response element, lansoprazole-induced HO-1 reporter gene activity was not increased. Taken together with our results, lansoprazole or pantoprazole up-regulated HO-1 expression and this up-regulated HO-1 contributed to the anti-inflammatory effects. Against gastric damage induced by NSAIDs, PPIs significantly reduced the mRNA expression and production of TNF-α and IL-1β in THP-1 cells stimulated by other irritants such as lipopolysaccharide (LPS) and
H. pylori water extracts [
15]. Lansoprazole inhibited the phosphorylation and degradation of inhibitory factor kappaB-alpha (IκBα) and phosphorylation of ERK in THP-1 cells, reaching to the conclusion that PPIs could exert anti-inflammatory effects by directly suppressing induction of TNF-α and IL-1β via the inhibition of NF-κB and ERK activation in inflammatory cells. Though they used inflammatory cells and we used gastric mucosal cells, we observed similar results that PPIs-induced HO-1 significantly inhibited indomethacin-induced levels of TNF-α and IL-1β. Natale G
et al.[
14] have reported that pretreatment with 90 μM/kg lansoprazole significantly prevented alcohol-induced gastric damage, suggesting a significant reduction of gastric oxidative stress associated with an increased bioavailability of mucosal sulfhydryl (SH) compounds.
Then, the question arises whether there might be difference in PPIs-induced protective mechanisms according to kinds of NSAIDs or kinds of PPIs. Blandizzi C
et al.[
29] treated male Sprague–Dawley rats with several kinds of NSAIDs, 100 μM/kg indomethacin, 60 μM/kg diclofenac, 150 μM/kg piroxicam or 150 μM/kg ketoprofen. Thirty minutes before NSAIDs, animals were orally treated with lansoprazole and four hours after the end of treatment, gastric mucosal PGE
2, malondialdehyde (MDA), myeloperoxidase (MPO) or non-protein sulfhydryl compounds (GSH) levels were measured, respectively. As result, PPIs prevented against NSAIDs-induced gastric damage irrespective of kinds of NSAIDs, mainly alleviating NSAIDs-induced mucosal oxidative injury [
28]. In preliminary study, we have also tested the HO-1 inducing capacity according to PPIs including lansoprazole, pantoprazole, rabeprazole, and omeprazole and we have found pantoprazole was the best in inducing HO-1 as well as other biological actions. We speculated these differences were based on the stability of PPI in aqueous condition.
Interestingly, these protective actions of PPIs against NSAIDs-induced gastric damages were not confined to just stomach. According to Takagi T
et al.[
27], they investigated whether PPIs ameliorated intestinal mucosal injuries induced by ischemia-reperfusion in rats and found that lansoprazole or pantoprazole have been demonstrated to prevent gastrointestinal mucosal injury by mechanisms independent of acid inhibition. Esomeprazole also counteracted the detrimental action of indomethacin on ulcer repair through both acid-dependent and acid-independent mechanisms [
26]. Though subtle difference in the action mechanisms related to acid suppression, PPIs irrespective of kinds were similar in the protective action beyond acid suppression, but we have used pantoprazole in our investigation, which shows utmost protective actions against NSAIDs-induced gastric damages based on HO-1 induction.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
KBH first set the hypothesis, mentored, and designed the whole experiments and my co-authors, who contributed to the completion of current manuscript, YMH as a PhD student and HJL, a professor in the lab of chemoprevention contributed to most experimental works. EHK and YJK contributed to complete the current study and manuscript. All authors read and approved the final manuscript.