Substance P released by TRPV1-expressing neurons produces reactive oxygen species that mediate ethanol-induced gastric injury

Free Radical Biology & Medicine 43 (2007) 581 – 589 www.elsevier.com/locate/freeradbiomed Original Contribution Substance P released by TRPV1-expressing neurons produces reactive oxygen species that mediate ethanol-induced gastric injury David Gazzieri a,1 , Marcello Trevisani a,1 , Jochen Springer b , Selena Harrison c , Graeme S. Cottrell d , Eunice Andre a , Paola Nicoletti c , Daniela Massi e , Sandra Zecchi f , Daniele Nosi f , Marco Santucci e , Norma P. Gerard g , Monica Lucattelli h , Giuseppe Lungarella h , Axel Fischer b , Eileen F. Grady d , Nigel W. Bunnett d , Pierangelo Geppetti a,c,⁎ a f Center of Excellence for the Study of Inflammation, University of Ferrara, Viale Pieraccini 6, 50139 Florence, Italy b Allergy Center Charité, Berlin, Germany c Department of Critical Care Medicine and Surgery, University of Ferrara, Viale Pieraccini 6, 50139 Florence, Italy d Departments of Surgery and Physiology, University of California at San Francisco, San Francisco, CA, USA e Department of Human Pathology and Oncology, University of Ferrara, Viale Pieraccini 6, 50139 Florence, Italy Department of Anatomy, Histology, and Forensic Medicine, University of Florence, University of Ferrara, Viale Pieraccini 6, 50139 Florence, Italy g Department of Pediatrics, Harvard Medical School, Cambridge, MA, USA h Department of Physiopathology, Experimental Medicine and Public Health, University of Siena, Siena, Italy Received 28 December 2006; revised 11 May 2007; accepted 11 May 2007 Available online 18 May 2007 Abstract Although neurokinin 1 receptor antagonists prevent ethanol (EtOH)-induced gastric lesions, the mechanisms by which EtOH releases substance P (SP) and SP damages the mucosa are unknown. We hypothesized that EtOH activates transient receptor potential vanilloid 1 (TRPV1) on sensory nerves to release SP, which stimulates epithelial neurokinin 1 receptors to generate damaging reactive oxygen species (ROS). SP release was assayed in the mouse stomach, ROS were detected using dichlorofluorescein diacetate, and neurokinin 1 receptors were localized by immunofluorescence. EtOH-induced SP release was prevented by TRPV1 antagonism. High dose EtOH caused lesions, and TRPV1 or neurokinin 1 receptor antagonism and neurokinin 1 receptor deletion inhibited lesion formation. Coadministration of low, innocuous doses of EtOH and SP caused lesions by a TRPV1-independent but neurokinin 1 receptor-dependent process. EtOH, capsaicin, and SP stimulated generation of ROS by superficial gastric epithelial cells expressing neurokinin 1 receptors by a neurokinin 1 receptor-dependent mechanism. ROS scavengers prevented lesions induced by a high EtOH dose or a low EtOH dose plus SP. Gastric lesions are caused by an initial detrimental effect of EtOH, which is damaging only if associated with TRPV1 activation, SP release from sensory nerves, stimulation of neurokinin 1 receptors on epithelial cells, and ROS generation. © 2007 Elsevier Inc. All rights reserved. Keywords: TRPV1; Reactive oxygen species; Substance P; Gastric lesions; Ethanol Introduction Abbreviations: EtOH, TRPV1, transient receptor potential vanilloid 1; SP, substance P; NK1R, neurokinin 1 receptor; CGRP, calcitonin gene-related peptide; ROS, reactive oxygen species; LI, like immunoreactivity; ig, intragastric; DCFDA, dichlorofluorescein diacetate; PBS, phosphate-buffered saline; RT-PCR, reverse transcriptase-polymerase chain reaction. ⁎ Corresponding author. Department of Critical Care Medicine and Surgery, Section of Geriatric Medicine and Urology, University of Florence, Viale Pieraccini 6, 50139 Florence, Italy. Fax: +39 55 4271280. E-mail address: pierangelo.geppetti@unifi.it (P. Geppetti). 1 These authors made equal contributions. 0891-5849/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.freeradbiomed.2007.05.018 Ingestion of elevated amounts of ethanol (EtOH) can cause hemorrhagic ulceration of the stomach in humans and experimental animals by mechanisms that are incompletely understood [1,2]. Although a direct toxic action of EtOH is likely, independent factors such as gastric acidity and submucosal vasodilatation are important determinants of lesion severity [3,4]. EtOH also stimulates generation of mediators (e.g., leukotrienes, prostanoids, mast cell products, autacoids) that contribute to its detrimental effects [1–4]. 582 D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 Nerve fibers derived from polymodal sensory neurons of dorsal root and vagal ganglia innervate the gastric mucosa. Some of these neurons are sensitive to the vanilloid capsaicin, the pungent ingredient of hot peppers, and express the calcitonin gene-related peptide (CGRP), and the tachykinins substance P (SP) and neurokinin A (NKA). Sensitivity to capsaicin depends on expression of transient receptor potential vanilloid 1 (TRPV1) [5], a nonselective cation channel activated by noxious heat (43–52°C) [5], protons [6,7], and certain lipids [8,9]. TRPV1 activation results in the influx of cations into nerve terminals and the release of CGRP, SP, and NKA. In many tissues, including the stomach, CGRP causes vasodilatation, and SP/NKA induce plasma extravasation and granulocyte infiltration (neurogenic inflammation) [10]. Calcitonin receptor like receptor and receptor activity modifying protein 1 compose the CGRP receptor [11], whereas the neurokinin 1 receptor (NK1R) mediates the effects SP/NKA [12]. CGRP, released from gastric sensory nerve terminals, strongly stimulates gastric blood flow and thereby protects against injury [3,4]. However, the role of the tachykinins in gastric injury is less well understood. SP potentiates gastric lesions induced by EtOH [13], and the SP antagonists, [D-Pro2, D-Trp7,8]SP and spantide, markedly reduce EtOH-induced gastric lesions [13,14]. Various mechanisms have been proposed to mediate this detrimental action of SP, including vasoconstriction and mast cell activation with subsequent release of leukotrienes or platelet-activating factor [13,14]. However, none of these proposals has been satisfactorily demonstrated and the mechanism of SP-mediated gastric injury is unknown. We recently reported that EtOH activates TRPV1 [15], probably by reducing the threshold temperature (43°C) [5] required for its activation. This effect of EtOH, which explains the sensation of burning pain when alcohol is applied to wounds, results in the release of neuropeptides and consequent neurogenic inflammation in the esophagus [15] and airways [16]. However, the contribution of this mechanism to alcoholinduced gastric ulceration is unknown. We investigated the hypothesis that EtOH activates TRPV1 to release SP from sensory nerve terminals in the gastric mucosa, and that SP activates the NK1R on gastric epithelial cells to exacerbate the harmful effects of alcohol and produce hemorrhagic gastric lesions. We also investigated the mechanism by which SP contributes to these lesions. Materials and methods Animals Male Sprague-Dawley rats (∼250 g), BALB/c, and C57BL/ 6J mice (∼ 25 g) were from Charles River Laboratories (Milan, Italy) and Morini (Reggio Emilia, Italy). NK1R−/− mice were backcrossed to BALB/c and C57BL/6J backgrounds [17]. The Ethical Committee for Animal University of Florence and the Institutional Animal Care and Use Committee of UCSF approved all experiments. Animals were killed using sodium pentobarbitone (200 mg/kg ip). CGRP and SP release Slices of the oxyntic region of the mouse stomach (∼ 100 mg) were placed in 2-ml chambers and superfused at 0.4 ml/min with Krebs solution [15] containing D-glucose 11 mM, with 0.1% BSA, 1 μM phosphoramidon, and 1 μM captopril and gassed (95% O2, 5% CO2) at 37°C. Tissues were pretreated (for 20 min) with capsaicin (10 μM), capsazepine (10 μM), Ca2+-free medium with 1 mM EDTA, or vehicle prior to stimulation with 3% EtOH. Perfusate was collected at 10-min intervals (2 prestimulus, 2 stimulus, 1 poststimulus), and analyzed by enzyme immunoassays for CGRP and SP [18]. Detection limits were 5 pg/ml for CGRP and 2 pg/ml for SP. Peptide release was calculated by subtracting the mean prestimulus value from values obtained during and after stimulation (expressed as fmol of peptide/g wet weight tissue/ 20 min). Gastric lesions Mice and rats were fasted overnight. BALB/c mice received vehicle (0.9% NaCl), 60% EtOH (high dose), or 30% EtOH (low dose) (23 ml/kg) by gavage. Because in preliminary experiments C57/BL6J mice were found to be slightly more sensitive to EtOH if compared to BALB/c mice, a dose of 40% EtOH, that produced lesions that covered ∼ 50% of the glandular surface, was used. Rats received vehicle, 90% EtOH (high dose), or 45% EtOH (low dose) (14 ml/kg) by gavage. Mice and rats were killed 10, 30, or 60 min later. To determine the area of gastric lesions, the stomach was opened, photographed, and gross gastric injury was assessed by computerized planimetry. The area of mucosal hemorrhagic damage was expressed as a percentage of the total area of the glandular mucosa [19]. Histology and immunohistochemistry For histopathological analysis, the oxyntic region of the stomach was fixed in 10% buffered formalin and embedded in paraffin, and 5-μm sections were stained with hematoxylin and eosin. Specimens were examined for histopathology in a blinded fashion. For immunohistochemistry 4-μm sections were dewaxed in Bio-Clear (Bio-Optica, Milan, Italy) and hydrated with graded EtOH concentrations. Antigen retrieval was routinely performed by immersing the slides in a thermostated bath containing 10 mM citrate buffer (pH 6.0) for 15 min at 97°C followed by cooling for 20 min at room temperature. Endogenous peroxidase activity was blocked with hydrogen peroxide at 3% in distilled water for 10 min. After blocking with normal horse serum (UltraVision, LabVision, Fremont, CA), sections were incubated with rabbit polyclonal anti 4-hydroxytrans-2,3-nonenal (HNE) antiserum (4-hydroxy-2-nonenal, Alpha Diagnostic, San Antonio, TX) at 1:500 dilution for 30 min. Staining was achieved using an EnVision detection system, peroxidase/DAB (Dako, Glostrup, DK) for 30 min at room temperature. Signal was detected using 3,3′-diaminobenzidine (Dako) as chromogen. Nuclei were slightly counter- D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 stained with Mayer's hematoxylin. Negative control was performed by substituting the primary antibody with a nonimmune mouse serum. ROS assay Mice and rats were anesthetized with ketamine hydrochloride and xylazine hydrochloride (both 10 mg/kg ip). ROS production was assayed in the oxyntic mucosa as described for mouse lung, with slight modifications [20]. Dichlorofluorescein diacetate (DCFDA, 1 μM, Alexsis, Grünberg, Germany) was administered by gavage with the aforementioned stimuli (vehicle, EtOH, capsaicin, and SP). After 30 min, animals were killed and 4% paraformaldehyde was injected into the stomach (20 min, 4°C). The stomach was washed with PBS, and the oxyntic region was mounted in carbonatebuffered glycerol. ROS levels were determined by assay of DCFDA, which is transformed by oxygen radicals and can be detected at 488 nm by using a confocal microscope. At least 30 crypts per animal were analyzed in a blind fashion. In some experiments, z-projections were taken through the mucosa at 2.1-μm intervals. RT-PCR Total RNA (1 μg) prepared from dissected oxyntic mucosa of rat and mouse stomach was reverse-transcribed and amplified by PCR using primers specific for mouse (forward, 5′-gtgcaacctacctggcaaat-3′; reverse, 5′-accagcagaggcaggaagta3′) and rat (forward, 5′-tacttcctgcctctgctggt-3; reverse, 5′tgaccttgtacacgctgctc-3′) NK1R. Controls omitted RT. Products were separated using a 2% agarose gel, stained with ethidium bromide, and sequenced. Immunofluorescence Mice and rats were transcardially perfused with 4% paraformaldehyde in 100 mM PBS, pH 7.4, and the oxyntic region of the stomach was placed in paraformaldehyde (12 h, 4°C). Frozen sections (16 μm) were prepared. Sections were incubated in PBS containing 5% NGS and 0.3% Triton for 30 min, and incubated with primary antibodies: rabbit anti-rat NK1R (1:1000) [21], guinea pig anti-SP (1:1000, Chemicon International), or NK1R antibody preabsorbed with the receptor fragment used for immunization (10 μM, 24 h, 4°C) for 16 h at 4°C. Tissue was washed and incubated with FITC-conjugated goat anti-rabbit IgG and Rhodamine Red X-conjugated goat anti-guinea pig IgG (Jackson ImmunoResearch, West Grove, PA) (1:200, 2 h, room temperature). Tissues were examined by confocal microscopy [22]. Drug treatments Mice were treated with NK1R antagonist SR140333 (1.6 μmol/kg, iv; gift from Dr X. Emonds-Alt, Sanofi Recherché, Montpellier), TRPV1 antagonist capsazepine (10 μmol/kg, sc), CGRP receptor antagonist CGRP(8–37) (0.4 μmol/kg iv), or 583 scavengers of reactive oxygen species (ROS) N-acetylcysteine (1200 mg/kg, intragastric, ig), ascorbic acid (1200 mg/kg, ig), lipolic acid (50 mg/kg, ip), or vehicles. Intravenous, subcutaneous, and intragastric administrations were at 15, 60, and 90 min prior to the EtOH, respectively. SP (1 μmol/kg, iv) or vehicle was administered immediately before EtOH. Capsaicin (10 mg/kg) was administered by gavage. Statistical analysis Data are expressed as mean ± SE, and compared using Student's t test, ANOVA, and Dunnett's test, or Kruskal-Wallis and the Mann-Whitney U test, with P b 0.05 considered significant. Results EtOH stimulated release of gastric sensory neuropeptides EtOH (3% v/v in 0.9% NaCl) stimulated release of SP-like immunoreactivity (LI) and CGRP-LI from slices of the oxyntic region of the mouse stomach (Figs. 1a and b). Preexposure of tissue to capsaicin (10 μM, 20 min) to desensitize TRPV1 abolished this response. Depletion of extracellular Ca2+ ions (Ca2+ -free medium plus 1 mM EDTA) and the TRPV1 antagonist capsazepine (10 μM) also prevented the stimulatory effect of EtOH. Thus, EtOH induces, via TRPV1 activation, a Ca2+-dependent release of SP and CGRP from capsaicinsensitive sensory neurons. EtOH induced gastric lesions by TRPV1- and SP-mediated mechanisms Intragastric administration of EtOH vehicle (0.9% NaCl) to BALB/c mice did not cause visible lesions of the oxyntic mucosa (Figs. 2 and 3). However, a high dose of EtOH (60% v/v, 0.9% NaCl, ig) caused a large area of diffuse reddening of the mucosa within 10 min (not shown), which was moderately increased at 30 min and maintained at 60 min (the 30-min time point was selected for further study) (Figs. 2 and 3a). Pretreatment with the TRPV1 antagonist, capsazepine, the NK 1 R antagonist, SR140333, or the ROS scavengers, N-acetylcysteine, lipolic acid, or ascorbic acid, strongly inhibited the effects of high dose EtOH on lesion formation (Figs. 2 and 3a). In contrast, the CGRP receptor antagonist, CGRP(8–37), moderately increased the area of EtOH-induced gastric lesions (Fig. 3a). The effects of high dose EtOH were markedly reduced in NK1R−/− mice (BALB/c or C57/BL6J backgrounds) compared to wild-type mice (Fig. 3c). However, capsazepine, SR140333, and Nacetylcysteine were similarly effective in reducing lesions caused by high dose EtOH in wild-type C57/BL6J mice (data not shown) and BALB/c mice. When administered separately, a low dose of EtOH (30% ig) or SP (1 μmol/kg iv) did not cause detectable lesions in BALB/c mice, and SR140333 did not alter this effect of low dose EtOH (Figs. 2 and 3b). However, when coadministered, low dose EtOH and SP increased the lesion area (Figs. 2 and 3b). 584 D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 Fig. 1. Effect of 3% EtOH on release of SP (a) and CGRP (b) from mouse stomach in Ca2+-free medium or with capsaicin desensitization (CAP, n = 4), capsazepine (CPZ, n = 4), or vehicle (Veh, n = 6). *P b 0.01 vs Veh. SR140333 and N-acetylcysteine, but not capsazepine, prevented formation of these lesions (Fig. 3b). Identical results were obtained in rats (Figs. 3a and b). Lesions induced by high dose EtOH were moderately increased by the CGRP receptor antagonism and inhibited by TRPV1 and NK1R antagonism and by ROS scavengers. Coadministration of a low dose EtOH and SP caused lesions in rats, and SR140333 or ROS scavengers, but not capsazepine, prevented these lesions (Fig. 3b). Fig. 2. Macroscopic (inset, luminal surface) and microscopic (section through gastric wall) images of the mouse stomach at 30 min after exposure to high dose (HD) or low dose (LD) EtOH or vehicle. Capsazepine (CPZ), SR140333 (SR), or N-acetylcysteine (NAC) inhibited lesion formation induced by high dose EtOH (arrows). Low dose EtOH or SP alone caused minimal damage, but together caused lesion formation (arrows). D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 585 Fig. 3. Effects of high dose (HD) or low dose (LD) ethanol on gastric lesion formation in mice and rats. (a) Capsazepine (CPZ, n = 8 and n = 8, respectively), SR140333 (SR, n = 8 and n = 8, respectively), N-acetylcysteine (NAC, n = 8 and n = 6, respectively), lipolic acid (LA, n = 8 and n = 6, respectively), and ascorbic acid (AA, n = 8 and n = 6, respectively) prevented the lesions to high dose EtOH (Veh, n = 8 and n = 8, respectively), whereas CGRP(8–37) (n = 8 and n = 6, respectively) exacerbated the damage. (b) Lesions caused in mice and rats by a low dose ethanol plus SP (n = 8 and n = 8, respectively), prevention by SR140333 (n = 8 and n = 6, respectively), NAC (n = 8 and n = 6, respectively) but not CPZ (n = 8 and n = 6, respectively). (c) High dose ethanol caused lesions in wild-type (WT) but not NK1R knockout (KO) mice of both strains (n = 8 in each condition). Control mice and rats received 0.9% NaCl (Con, n = 6 and n = 6, respectively), with vehicles of SR140333 (Veh1, n = 6 and n = 6, respectively) or CPZ, SR140333, and NAC (Veh2, (n = 8 and n = 6, respectively). *P b 0.05 vs Veh2 or WT. #P b 0.05 vs Veh1. EtOH caused hemorrhagic lesions of the gastric mucosa Histological examination indicated that high dose EtOH caused severe damage of the oxyntic mucosa of the mouse stomach, consisting of acute erosive hemorrhagic lesions, with diffuse coagulative cell necrosis, multiple superficial erosions, marked vascular congestion, and extravasation of erythrocytes (Fig. 2). Scattered inflammatory cells, including neutrophils, were present within the deep mucosal and submucosal layers. A low dose of EtOH produced only a slight submucosal edema and ectatic blood vessels, with no extravasation of erythrocytes, and no significant cellular infiltrate or mucosal damage (Fig. 2). SP alone did not cause detectable histopathological change in the mucosa, but did induce dilation of submucosal blood vessels surrounded by focal areas of neutrophilic infiltration (Fig. 2). However, coadministration of low dose EtOH and SP induced damage similar to that obtained with high dose EtOH, consisting of marked edema in the submucosa, prominent vascular congestion in the mucosa and submucosa, and multiple foci of coagulative cell necrosis in the mucosa (Fig. 2). Capsazepine, SR140333, or N-acetylcysteine inhibited all features of damage and inflammation induced high dose EtOH (Fig. 2). SR140333 or N-acetylcysteine, but not capsazepine, inhibited damage and inflammation induced by low dose EtOH and SP (not shown). EtOH stimulated ROS generation by TRPV1- and SP-mediated mechanisms EtOH vehicle (0.9% NaCl) did not affect the generation of ROS in the mouse stomach, when assessed by DCFDA fluorescence (Figs. 4a and b). In contrast, high dose EtOH markedly increased ROS generation in the gastric mucosa. Capsazepine, SR140333, or N-acetylcysteine prevented EtOHinduced ROS generation. Capsaicin (10 mg/kg ig) also increased ROS generation in the gastric mucosa, and capsazepine or SR140333 abolished this effect (Fig. 4b). SP (1 μmol/kg, iv) also stimulated ROS generation, and SR140333, but not capsazepine, abolished this effect (Fig. 4b). In contrast, low dose EtOH did not stimulate ROS production (not shown). High dose EtOH, also stimulated ROS generation in the oxyntic mucosa of the rat stomach (Fig. 4c), an effect that was prevented by capsazepine, SR140333, and N-acetylcysteine (Fig. 4c). Examination of the sagittal view of the oxyntic mucosa of the mouse stomach for DCFDA-dependent fluorescence revealed that high dose EtOH induced ROS generation principally in the superficial layer of the epithelium (Fig. 5). Although some fluorescence was also detected in deeper layers under basal conditions, this was unaffected by EtOH. To add further support to the hypothesis that ROS generation was localized to the superficial part of the gastric mucosa immunohistochemistry for 586 D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 Fig. 4. Generation of ROS in the mouse and rat oxyntic mucosa determined by DCFDA-dependent fluorescence. (a) Pseudo-color images of ROS generation at the surface mucosa of the mouse stomach (red, high; blue, low). (b) ROS generation in the oxyntic mucosa of mouse (n = 8 in each condition) and rat (n = 6 in each condition) (c) stomach. Luminal high dose (HD) ethanol or capsaicin stimulated ROS generation, and capsazepine (CPZ), SR140333 (SR), and N-acetylcysteine (NAC) inhibited this ROS generation. Intravenous SP stimulated ROS generation in mouse stomach, and SR140333, NAC, but not CPZ inhibited this response. Drug vehicles (Veh) had no effect. *P b 0.05 vs vehicle or saline controls. HNE was performed. Oxidative stress produces reactive carbonyl species (RCS) principally by peroxidation of plasma membrane polyunsaturated fatty acids. HNE is a specific and stable RCS alkylating agent that reacts with proteins, generating various forms of adducts (cysteine, lysine, histidene residues) [23]. Thus, HNE localization by immonohistochemistry is used as a reliable marker of ROS generation [24]. The moderate HNE immunostaining observed in the inner part of the mouse gastric mucosa after vehicle was slightly increased after administration of a high dose of EtOH (Fig. 5). In contrast, in the superficial part of the gastric mucosa HNE staining, that was practically absent after vehicle, was remarkably increased after the administration of a high EtOH dose (Fig. 5). NK1R and SP were expressed in the gastric mucosa Products of the anticipated size were amplified by RT-PCR from the mucosa of the mouse (557 bp) and rat (442 bp) oxyntic mucosa and identified by sequencing (Fig. 6a). Low levels of NK1R-LI were detected in epithelial cells of the gastric secretory mucosa of mice and rats (Fig. 6b). NK1R-LI was prominently detected in neurons of the myenteric plexus and in cells in the interstitial cells of Cajal (not shown). NK1R-LI was not detected in the epithelium when the NK1R antibody was preabsorbed with the receptor fragment used for immunization. SP was present in nerve fibers in the lamina propria of the epithelium in close proximity to epithelial cells expressing Fig. 5. Sagittal views of DFCDA-dependent fluorescence indicative of ROS generation (a and b, left panels) or immunohistochemistry for, 4-hroxy-trans-nonenal (a and b, right panels) of mouse stomach wall following an intragastric dose of ethanol (EtOH HD) (b) or its vehicle (a). D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 587 We observed that antagonism of CGRP receptor enhanced the injurious effect of high dose ethanol, confirming previous reports of a protective role of CGRP in the stomach [3,4]. This protection probably depends on CGRP-induced vasodilatation of submucosal arterioles, which results in clearance of backdiffused acid [3]. More importantly, antagonism or deletion of the NK1R prevented EtOH-induced gastric lesions. These results confirm previous findings obtained with first generation NK1R antagonists [13,14]. The protective effect of NK1R deletion excludes the possibility that antagonists are protective through nonspecific mechanisms. However, despite the opposing anti- and pro-ulcerogenic effects of CGRP and SP, our observation that EtOH causes damage by an NK1R-mediated mechanism indicates that the pro-ulcerogenic actions of tachykinins prevail. Contribution of TRPV1 to EtOH-induced gastric ulceration Fig. 6. (a) RT-PCR of NK1R expression in oxyntic mucosa of mouse and rat stomach. (b) Localization of NK1R and SP in oxyntic mucosa of rat and mouse stomach. Control shows preabsorption of NK1R antibody with receptor fragment. Scale = 20 μm. NK1R-LI (Fig. 6b), as well as in fibers in the myenteric plexus and muscularis externa (not shown). Discussion Our results show that EtOH activates TRPV1 on sensory nerve endings in the gastric mucosa to stimulate release of SP. High dose EtOH causes hemorrhagic lesions of the superficial gastric mucosa at sites of NK1R expression and ROS generation by a TRPV1- and NK1R-dependent mechanism. Moreover, SP potentiates the injurious effects of low dose EtOH. NK1R stimulates generation of ROS by the gastric epithelium, resulting in tissue damage. This mechanism is identical in mice and rats, and may be conserved among species. Thus, antagonism of TRPV1 and the NK1R may protect against the damaging effects of EtOH on the gastric mucosa, and possibly other epithelial tissues. Because of the essential role of TRPV1 in EtOH-induced SP release, and the reported contribution of SP to EtOH-induced gastric lesions [13,14], we hypothesized that a TRPV1 antagonist would not only block SP release but also reduce the lesions evoked by EtOH. Capsazepine markedly reduced EtOH-induced lesions, suggesting that EtOH-induced activation of TRPV1 is the first step of a cascade of events that result in gastric damage (Fig. 7, step 1). Protons can activate TRPV1 [6,7], and EtOH sensitizes proton-induced activation of TRPV1 [15]. Thus, after an initial EtOH-induced activation of TRPV1 to induce mucosal damage and backdiffusion of acid, protons and EtOH, may synergize to produce enhanced TRPV1 stimulation and massive release of sensory neuropeptides (Fig. 7, step 2). Further investigations are, however, required to define the contribution of hydrogen ions to EtOH-induced activation of TRPV1, SP release, and gastric damage. The primary role of SP and the NK1R in EtOH-induced gastric Anti- and pro-ulcerogenic effects of gastric sensory neuropeptides EtOH stimulated release of both SP and CGRP from the stomach, and capsaicin desensitization, removal of extracellular Ca2+, and capsazepine prevented this stimulation. Thus, EtOH activates TRPV1 on sensory nerve endings to stimulate neurosecretion of SP and CGRP, as observed in the esophagus and airways [15,16]. Perfusion of the gastric lumen with EtOH also promotes SP release [14]. However, CGRP and tachykinins have opposing anti- and pro-ulcerogenic effects, respectively. Fig. 7. Schematic representation of the proposed mechanism by which SP contributes to the ethanol-induced gastric hemorrhagic lesion in the mouse stomach. Ethanol after a still undetermined initial (1) action, by itself, or in combination with backdiffused acid stimulates TRPV1 to release SP (2) that by activation of epithelial NK1 receptors (3) generates cytotoxic reactive oxygen species (ROS) (4). The inhibitory effects of capsazepine (CPZ), the NK1 receptor antagonist, SR140333, and the ROS scavengers, N-acetylcysteine (NAC), lipolic acid (LA), and ascorbic acid (AA) on their respective targets are also reported. 588 D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 injury [13,14] can now be explained by the ability of EtOH to activate or sensitize TRPV1 to release tachykinins. Contribution of ROS to EtOH-induced gastric ulceration Our findings that SP and the NK1R are required for EtOHinduced gastric injury raised the question of the mechanism of NK1R-dependent tissue damage. Our observation that SP activates the NK1R in the mouse lung to increase ROS formation in epithelial cells [20] and induce expression of the pro-inflammatory AP-1 transcription factor [25] suggested a role for ROS in gastric damage. Indeed, ROS generation within the rat gastric mucosa is a major contributing factor to gastric lesions induced by EtOH, aspirin, and stress [26,27]. We observed that three structurally distinct ROS scavengers (Nacetylcysteine, lipolic acid, ascorbic acid) protected against EtOH-induced lesions. In the present work several observations suggest that EtOH increases ROS generation in the gastric mucosa via TRPV1- and NK1R-dependent mechanisms. First, luminal capsaicin stimulated ROS generation and capsazepine and SR140333 prevented this effect. Thus, activation of TRPV1 stimulates ROS generation by a SP- and NK1R-dependent process. Second, SP stimulated ROS formation, which was abolished by SR140333 and thus mediated by the NK1R. Finally, EtOH strongly stimulated ROS generation, which was abolished by capsazepine and SR140333, and thus dependent on TRPV1 stimulation, SP release, and NK1R activation. EtOH-induced ROS formation was confined to the superficial layers of the gastric epithelium. A cause and effect relationship between ROS production and generation of gastric lesions is supported by the observation that distinct pharmacological interventions that abolished ROS formation (capsazepine, SR140333, ROS scavengers) also prevented gastric damage. Protective agents usually do not prevent EtOH-induced damage of the superficial epithelium. For example, capsaicin, presumably via the CGRP-induced vasodilatation, does not prevent EtOH from causing superficial damage, and solely protects against deep mucosal lesions [28]. Prostaglandins also do not reduce damage to surface epithelial cells of the stomach [29]. Thus, EtOH-induced lesions at the surface of the gastric mucosa result from a specific mechanism exclusively sensitive to decreased ROS production. The requirement of ROS for EtOH-induced gastric lesions suggests that EtOH per se is the essential initial step (Fig. 7, step 1), but is not sufficient to develop the injury, because gastric lesions are produced only if SP is released (Fig. 7, step 2), and NK1R activated (Fig. 7, step 3). This hypothesis is consistent with previous findings that NK1R agonists (SP, septide, senktide) potentiate EtOH-induced gastric lesions [13]. Thus, we hypothesized that a low dose of EtOH that per se produces no or negligible gastric lesion can induce a marked damage only when coadministered with a dose of SP capable of generating ROS. This hypothesis was confirmed in the present work, and ROS generation is essential for epithelial damage (Fig. 7, step 4), since a ROS scavenger inhibited the injury induced by EtOH plus SP. However, ROS generation per se is not sufficient to cause gastric lesions because capsaicin and SP strongly stimulated ROS generation without detectable tissue damage. Therefore, an additional action of EtOH is required. The nature of this essential effect of EtOH remains to be determined. Although a major role for ROS in EtOH-induced gastric lesions has been proposed [26,27], the source of ROS is controversial. Neutrophils are a major source of ROS in some damaged tissues. However, in our study, EtOH and SP induced influx of only small numbers of neutrophils, mostly around submucosal vessels, and neutrophils were rare or absent from superficial layers of the gastric mucosa where ROS production and necrosis occurred. Moreover, there is no evidence that neutrophils express functional NK1Rs. Thus, neutrophils are unlikely to be the source of ROS in the stomach after EtOH or SP exposure. SP-induced damage of the gastric mucosa has been ascribed to the degranulation of mast cells, an effect that is inhibited by ketotifen [13]. However, amphiphilic peptides, such as SP, release mediators from mast cells only at very high concentrations (N μM) by a nonreceptor-mediated mechanism [30]. Our observations that antagonism or deletion of the NK1R prevented EtOH-evoked lesions and ROS formation, together with the absence of functional NK1Rs in mast cells, suggest that mast cells do not contribute to EtOH-evoked gastric damage. Moreover, ketotifen may reduce ROS formation, as observed in primed eosinophils [31]. We observed that EtOH stimulated ROS formation in the superficial area of the gastric mucosa and this observation was confirmed by immunohistochemistry for HNE, a stable product of oxidative stressinduced lipid peroxidation of plasma membrane polyunsaturated fatty acids [23]. In fact, EtOH produced a dramatic increase in HNE staining selectively in the superficial part of the gastric mucosa. These findings further exclude the role of neutrophils in the generation ROS by EtOH, because in the superficial part of the gastric mucosa, where ROS and HNE expression was markedly upregulated, no neutrophil infiltration was observed, as assessed by histological examination. Detection of NK1R-LI and of NK1R mRNA in gastric epithelial cells in close proximity to SP-containing nerve fibers supports the view that SP released from sensory nerve terminals may activate NK1R to generate ROS. The precise identity of the NK1R-expressing cells remains to be determined, although NK1R has been detected in Chief cells in the oxyntic mucosa of the dog stomach [32]. Thus, by analogy with the observations in the airway epithelium [20], we propose that SP activates the NK1R on gastric epithelial cells to promote the generation of cytotoxic ROS, thereby causing hemorrhagic lesions of the stomach wall. Acknowledgments This work was supported by MIUR, Rome (PRIN 2003054380), and NIH (DK39957, DK43207, DK57840, DK52388). References [1] Teyssen, S.; Singer, M. V. Alcohol-related diseases of the oesophagus and stomach. Best Pract. Res. Clin. Gastroenterol. 17:557–573; 2003. [2] Siegmund, S.; Haas, S.; Schneider, A.; Singer, M. V. Animal models in D. Gazzieri et al. / Free Radical Biology & Medicine 43 (2007) 581–589 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] gastrointestinal alcohol research—A short appraisal of the different models and their results. Best Pract. Res. Clin. Gastroenterol. 17:519–542; 2003. Holzer, P. Neural emergency system in the stomach. Gastroenterology 114:823–839; 1998. Holzer, P. Efferent-like roles of afferent neurons in the gut: blood flow regulation and tissue protection. Auton. Neurosci. 125:70–75; 2006. Caterina, M. J.; Schumacher, M. A.; Tominaga, M.; Rosen, T. A.; Levine, J. D.; Julius, D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389:816–824; 1997. Bevan, S.; Geppetti, P. Protons: small stimulants of capsaicin-sensitive sensory nerves. Trends Neurosci. 17:509–512; 1994. Tominaga, M.; Caterina, M. J.; Malmberg, A. B.; Rosen, T. A.; Gilbert, H.; Skinner, K.; Raumann, B. E.; Basbaum, A. I.; Julius, D. The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron 21:531–543; 1998. Huang, S. M.; Bisogno, T.; Trevisani, M.; Al-Hayani, A.; De Petrocellis, L.; Fezza, F.; Tognetto, M.; Petros, T. J.; Krey, J. F.; Chu, C. J.; Miller, J. D.; Davies, S. N.; Geppetti, P.; Walker, J. M.; Di Marzo, V. An endogenous capsaicin-like substance with high potency at recombinant and native vanilloid VR1 receptors. Proc. Natl. Acad. Sci. U. S. A. 99:8400–8405; 2002. Zygmunt, P. M.; Petersson, J.; Andersson, D. A.; Chuang, H.; Sorgard, M.; Di Marzo, V.; Julius, D.; Hogestatt, E. D. Vanilloid receptors on sensory nerves mediate the vasodilator action of anandamide. Nature 400: 452–457; 1999. Geppetti, P.; Holzer, P. Neurogenic Inflammation. CRC Press, Boca Raton, FL; 1996. McLatchie, L. M.; Fraser, N. J.; Main, M. J.; Wise, A.; Brown, J.; Thompson, N.; Solari, R.; Lee, M. G.; Foord, S. M. RAMPs regulate the transport and ligand specificity of the calcitonin-receptor-like receptor. Nature 393:333–399; 1998. Regoli, D.; Boudon, A.; Fauchere, J.-L. Receptors and antagonists for substance P and related peptides. Pharmacol. Rev. 46:551–599; 1994. Karmeli, F.; Eliakim, R.; Okon, E.; Rachmilewitz, D. Gastric mucosal damage by ethanol is mediated by substance P and prevented by ketotifen, a mast cell stabilizer. Gastroenterology 100:1206–1216; 1991. Hayashi, H.; Nishiyama, K.; Majima, M.; Katori, M.; Saigenji, K. Role of endogenous substance P in ethanol-induced mucosal damage in the rat stomach. J. Gastroenterol. 31:314–322; 1996. Trevisani, M.; Smart, D.; Gunthorpe, M. J.; Tognetto, M.; Barbieri, M.; Campi, B.; Amadesi, S.; Gray, J.; Jerman, J. C.; Brough, S. J.; Owen, D.; Smith, G. D.; Randall, A. D.; Harrison, S.; Bianchi, A.; Davis, J. B.; Geppetti, P. Ethanol elicits and potentiates nociceptor responses via the vanilloid receptor-1. Nat. Neurosci. 5:546–551; 2002. Trevisani, M.; Gazzieri, D.; Benvenuti, F.; Campi, B.; Dinh, Q. T.; Groneberg, D. A.; Rigoni, M.; Emonds-Alt, X.; Creminon, C.; Fischer, A.; Geppetti, P.; Harrison, S. Ethanol causes inflammation in the airways by a neurogenic and TRPV1-dependent mechanism. J. Pharmacol. Exp. Ther. 309:1167–1173; 2004. Bozic, C. R.; Lu, B.; Hopken, U. E.; Gerard, C.; Gerard, N. P. Neurogenic amplification of immune complex inflammation. Science 273:1722–1725; 1996. Frobert, Y.; Nevers, M. C.; Amadesi, S.; Volland, H.; Brune, P.; Geppetti, P.; Grassi, J.; Creminon, C. A sensitive sandwich enzyme immunoassay for [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] 589 calcitonin gene-related peptide (CGRP): characterization and application. Peptides 20:275–284; 1999. Wultsch, T.; Painsipp, E.; Thoeringer, C. K.; Herzog, H.; Sperk, G.; Holzer, P. Endogenous neuropeptide Y depresses the afferent signaling of gastric acid challenge to the mouse brainstem via neuropeptide Y type Y2 and Y4 receptors. Neuroscience 136:1097–1107; 2005. Springer, J.; Fischer, A. Substance P-induced pulmonary vascular remodelling in precision cut lung slices. Eur. Respir. J. 22:596–601; 2003. Grady, E. F.; Baluk, P.; Bohm, S.; Gamp, P. D.; Wong, H.; Payan, D. G.; Ansel, J.; Portbury, A. L.; Furness, J. B.; McDonald, D. M.; Bunnett, N. W. Characterization of antisera specific to NK1, NK2, and NK3 neurokinin receptors and their utilization to localize receptors in the rat gastrointestinal tract. J. Neurosci. 16:6975–6986; 1996. Cottrell, G. S.; Roosterman, D.; Marvizon, J. C.; Song, B.; Wick, E.; Pikios, S.; Wong, H.; Berthelier, C.; Tang, Y.; Sternini, C.; Bunnett, N. W.; Grady, E. F. Localization of calcitonin receptor-like receptor and receptor activity modifying protein 1 in enteric neurons, dorsal root ganglia, and the spinal cord of the rat. J. Comp. Neurol. 490:239–255; 2005. Esterbauer, H.; Schaur, R. J.; Zollner, H. Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic. Biol. Med. 11:81–128; 1991. Rahman, I.; van Schadewijk, A. A.; Crowther, A. J.; Hiemstra, P. S.; Stolk, J.; MacNee, W.; De Boer, W. I. 4-Hydroxy-2-nonenal, a specific lipid peroxidation product, is elevated in lungs of patients with chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med. 166: 490–495; 2002. Springer, J.; Pleimes, D.; Scholz, F. R.; Fischer, A. Substance P mediates AP-1 induction in A549 cells via reactive oxygen species. Regul. Pept. 124:99–103; 2005. Kwiecien, S.; Brzozowski, T.; Konturek, S. J. Effects of reactive oxygen species action on gastric mucosa in various models of mucosal injury. J. Physiol. Pharmacol. 53:39–50; 2002. Kwiecien, S.; Brzozowski, T.; Konturek, P. C.; Pawlik, M. W.; Pawlik, W. W.; Kwiecien, N.; Konturek, S. J. The role of reactive oxygen species and capsaicin-sensitive sensory nerves in the pathomechanisms of gastric ulcers induced by stress. J. Physiol. Pharmacol. 54:423–437; 2003. Holzer, P.; Pabst, M. A.; Lippe, I. T.; Peskar, B. M.; Peskar, B. A.; Livingston, E. H.; Guth, P. H. Afferent nerve-mediated protection against deep mucosal damage in the rat stomach. Gastroenterology 98:838–848; 1990. Szabo, S.; Trier, J. S.; Brown, A.; Schnoor, J. Early vascular injury and increased vascular permeability in gastric mucosal injury caused by ethanol in the rat. Gastroenterology 88:228–236; 1985. Mousli, M.; Bueb, J. L.; Bronner, C.; Rouot, B.; Landry, Y. G protein activation: a receptor-independent mode of action for cationic amphiphilic neuropeptides and venom peptides. Trends Pharmacol. Sci. 11:358–362; 1990. Yamada, Y.; Sannohe, S.; Saito, N.; Cui, C. H.; Ueki, S.; Oyamada, H.; Kanda, A.; Yamaguchi, K.; Hamada, K.; Adachi, T.; Kayaba, H.; Chihara, J. Effect of ketotifen on the production of reactive oxygen species from human eosinophils primed by eotaxin. Pharmacology 69:138–141; 2003. Vigna, S. R.; Mantyh, C. R.; Soll, A. H.; Maggio, J. E.; Mantyh, P. W. Substance P receptors on canine chief cells: localization, characterization, and function. J. Neurosci. 9:2878–2886; 1989.