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.