Article Text
Abstract
Objective: The human gastro-oesophageal junction is exposed to abundant amounts of luminal reactive nitrogen oxide species (RNOS) derived from the enterosalivary re-circulation of dietary nitrate. The aim of this study is to investigate the direct effects of luminal RNOS on the adjacent gastric barrier function using an ex vivo chamber model.
Methods: A chamber model in which the rat gastric mucosal membrane was mounted between the two halves of a chamber was designed to simulate the microenvironment of the lumen and the adjacent mucosa of the gastro-oesophageal junction. On the mucosal side of the chamber, RNOS were generated by the acidification of physiological concentrations of sodium nitrite. The epithelial barrier function was evaluated by electrophysiological transmembrane resistance, and membrane permeability with [3H]mannitol flux. The expression of occludin was evaluated by immunohistochemistry and immunoblotting. Dinitrosyl dithiolato iron complex (DNIC) was also measured by means of electron paramagnetic resonance spectroscopy to confirm the diffusion of RNOS from the mucosal lumen into the mounted mucosa.
Results: The administration of acidified nitrite to the mucosal lumen caused both a decrease in transmembrane resistance and an increase in epithelial permeability, suggesting a disturbance of the gastric barrier function. These changes were accompanied by a derangement of the expression of occludin. The diffusion of luminal RNOS into the mounted membrane was confirmed by showing the generation of DNIC within the tissue.
Conclusions: Simulating the microenvironment of the human gastro-oesophageal junction, this study demonstrated that RNOS generated luminally at the human gastro-oesophageal junction can derange the barrier function of the adjacent tissue by disrupting the tight junction.
Statistics from Altmetric.com
The incidence of adenocarcinoma of the cardia of the stomach and adjacent gastro-oesophageal (GO) junction has been increasing over the past 25 years, especially in Western countries.1 We recently reported that the characteristic features of this cancer, in contrast to ordinary non-cardiac gastric cancer, are that it preferentially occurs in subjects with preserved gastric acid secretion and is not associated with Helicobacter pylori infection.2 More recently, in relation to the rising incidence of this cancer, inflammation (carditis)3–9 and intestinal metaplasia9–13 localised to immediately below the human GO junction have received much attention. Since these histological findings are frequently observed even among those who are H pylori-negative,3 4 6 7 the causative factors for such histological events at the human GO junction remain obscure and they may well represent the histological response to many different types of insult.
After the ingestion of a high nitrate meal, a high level of salivary nitrite is sustained over several hours through the enterosalivary re-circulation of the dietary nitrate.14–17 A series of recent studies have demonstrated that when the nitrite in swallowed saliva encounters the acidic gastric environment at the human GO junction, it is immediately converted to nitrous acid (HNO2) and subsequently decomposes to a variety of reactive nitrogen oxide species (RNOS) including nitric oxide (NO), nitrogen dioxide (NO2), dinitrogen trioxide (N2O3) and dinitrogen tetroxide (N2O4).18–23 In the presence of sufficient amounts of ascorbic acid, which is secreted in human gastric acid, HNO2 is predominantly converted to NO.18–22 Subsequently, since these RNOS, in particular NO, are highly diffusible,24 they will be able to enter the adjacent tissue rapidly. Using an in vivo animal model, we recently demonstrated by means of electron paramagnetic resonance (EPR) spectroscopy that the NO generated luminally diffused into adjacent tissues to a substantial degree and at a level comparable with that of inducible nitric oxide synthase- (iNOS) derived NO production.25 Therefore, the human GO junction is likely to be a region of high nitrosative stress. However, whether the RNOS generated luminally at the GO junction affect the cell biology in the adjacent tissue remains to be clarified.
NO has been known to play various roles in the physiology and pathophysiology of the gastrointestinal tract, depending on its level. Although some studies demonstrated that, at lower levels, NO derived from constitutive nitric oxide synthase (cNOS) appears to have some beneficial effects on the gastrointestinal tract,26–28 other studies using lipopolysaccharide-treated animals have demonstrated that high concentrations of NO derived from iNOS can impair the intestinal barrier function by disrupting the tight junction.29–31 The highest concentrations of NO occurring in the body are not the result of enzymatic synthesis but rather result from chemical reactions within the lumen of the stomach, especially in the most proximal part of it, as described above.19 In this circumstance, once the NO and other RNOS diffuse from the acidic lumen into the tissue at neutral pH, they will be irreversibly metabolised to nitrite and nitrate. Therefore, it is the very surface of the epithelium of the GO junction that is exposed to the highest concentration of the RNOS, in contrast to examples of inflammation in human in which the immediate vicinity of inflammatory cells expressing iNOS would be exposed to the highest levels of endogenous RNOS. Hence, the tight junction of the surface epithelium at the GO junction is likely to be a principal target of the RNOS arising from the lumen.
In this study, we investigated the direct effects of luminal RNOS on the adjacent gastric barrier function using an ex vivo Ussing chamber model, in which the gastric mucosal membrane was mounted between the two halves of a chamber to simulate a microenvironment of the lumen and its adjacent mucosa of the human GO junction.
MATERIALS AND METHODS
Animals
Specific pathogen-free male Wistar rats, weighing approximately 250 g (CLEA Japan Inc., Tokyo, Japan) were used in all studies. Rats were deprived of food for 24 h but were allowed free access to water before the experiments. The approval of the Animal Welfare Committee of the Yamagata Promotional Organization for Industrial Technology and Tohoku University School of Medicine was obtained for all studies.
Ussing chamber system
The stomachs of the rats were rapidly removed after cervical dislocation under ether anaesthesia. Each stomach was opened up along the lesser curvature and washed in modified Krebs buffer containing Na+ 152 mM, K+ 2.5 mM, Ca2+ 2.5 mM, Mg2+ 1.2 mM, Cl– 136 mM, HCO3– 25 mM, PO43– 1.2 mM and glucose 11 mM to remove the luminal contents. The muscular layers were stripped off from the mucosa of the glandular stomach, and the remaining mucosa was mounted in an Ussing chamber with an exposed area of 1.47 cm2 (World Precision Instruments Inc., Sarasota, FL) as previously reported (fig 1).32 33 Other than the epithelial mucosal layer, the mounted mucosa comprised the underlying muscularis mucosa and a small amount of connective tissue. Two voltage-sensitive electrodes and two current-passing electrodes (World Precision Instruments Inc.) were connected to the two sides of the chamber via agar bridges. Both the mucosal and serosal sides of the chamber were connected to circulating reservoirs. The solution volume on each side was 15 ml, and was oxygenated and circulated with a 95% O2/5% CO2 gas-lift system. The reservoirs were water-jacketed to maintain the bathing solutions at 37°C.
Electrophysiological measurement
The transepithelial electrical potential difference (PD) in millivolts across the gastric mucosal membrane was measured directly. The transmembrane resistance in ohms×cm2 was determined using Ohm’s law by passing a 100 μA current through the membrane and by measuring the change in PD. The transmembrane resistance is considered to reflect the membranous integrity. Shunt resistance is one of the major components of the total transmembrane resistance and is known to be associated with the tight junction.34
Once the gastric mucosa was mounted in the Ussing chamber, both sides of the chamber were bathed with modified Krebs buffer for 60 min to ensure that the system was functioning and that the gastric mucosa was not damaged. The physiological basis of the transmucosal PD is believed to be a function of ionic flux at the mucosal level,35 and the PD can be used as an indicator of tissue viability. Hence, if the PD of the mounted mucosa was <10.0 mV, that preparation was excluded from this study. After this equilibration period, the mucosal side was perfused with pH 2.5 NaCl solution (154 mM), in which the pH was adjusted by 1.0 N HCl, plus physiological concentrations of sodium nitrite (0.1, 1.0 or 5.0 mM)36–38 in order to generate RNOS, or pH 2.5 NaCl alone as a control. Meanwhile, the serosal side was perfused with modified Krebs buffer at pH 7.4. The electrical measurements were made every 15 min over a period of 180 min, during which time the solutions on both sides of the chamber were exchanged after each measurement since the system is open to the air and the RNOS formed from nitrite in the acidic environment are assumed to decline substantially. In some experiments, the mucosal side was perfused with pH 7.4 Krebs buffer plus sodium nitrite (5 mM) in place of pH 2.5 solution, under which condition nitrite is known to be stable and few RNOS are generated.20 In addition, in another group of experiments, pH 2.5 NaCl plus sodium nitrate (5 mM) was administered in place of sodium nitrite; nitrate is known to be stable even at acidic pH. This sodium nitrate (5 mM) solution can be regarded as the same osmolarity control.
Measurement of membranous permeability
Mannitol is known to permeate the gastric membrane through a paracellular pathway,39 which is known to be formed by the tight junction. Hence, [3H]mannitol flux from the mucosal to the serosal side was measured to evaluate the permeability of the membrane. [3H]Mannitol (10 μCi; MP Biomedicals, Inc., Irvine, CA) was added to the mucosal side of the chamber and 1 ml aliquots were removed from the serosal side every 15 min. The aliquots from the serosal side of the chamber were placed into 5 ml of a cocktail solution (ACS-II; GE Healthcare Bio-Sciences Corp., Piscataway, NJ) for assessment in a liquid scintillation counter (BECKMAN LS-6500; Beckman Coulter Inc., Fullerton, CA).
Morphological analysis
At the end of the experiments, the mucosa was removed from the chamber, fixed in 10% buffered formalin and embedded in paraffin. Sections were mounted on glass slides and stained with H&E for light microscope analysis. Sections of untreated rat gastric mucosa, not mounted in the chamber, were made as the untreated control. Unstained sections were used for immunohistochemical analysis.
Immunohistochemistry for occludin
Occludin, an integral membrane protein involved in epithelial tight junction formation,40 was immunochemically stained for the gastric membrane after the chamber experiment. After deparaffinisation with xylene followed by a graded alcohol series, the sections were incubated with protease type XIV (Sigma, #P-5147) for protease pretreatment at 37°C for 10 min and then incubated in 3.0% H2O2 at room temperature for 10 min to block endogenous peroxidase activity. After washing with phosphate-buffered saline (PBS), the sections were blocked with 1.5% rabbit serum in the buffer for 30 min. Then, the sections were incubated overnight with rabbit anti-occludin polyclonal antibody (1:50) (Zymed Laboratories, San Francisco, CA). On the next day, the sections were washed with the buffer, and a VECTASTAIN Elite ABC rabbit IgG kit (Vector Laboratories, Burlingame, CA) was used according to the manufacturer’s instructions. The slides were counterstained with haematoxylin and then dehydrated with ethanol and xylene before permanent mounting and microscopic evaluation.
Immunoblotting for occludin
The expression of occludin was evaluated as described by Han et al.31 Mucosal specimens were homogenised in cold PBS and sonicated until the sample was completely dissolved. The samples were centrifuged at 12 000 g for 15 min at 4°C, and the supernatants were electrophoresed on 12% Bis-Tris Gel (Invitrogen, Carlsbad, CA).The proteins were electroblotted onto PVDF membranes (Invitrogen, California, USA) and were blocked with 5% skim milk for 60 min. The membranes were incubated overnight at 4°C with rabbit polyclonal anti-occludin antibody (1:100) (Zymed Laboratories). The immunoblots were exposed for 1 h to a 1:2000 dilution of the appropriate horseradish peroxidase-conjugated secondary antibody, and were impregnated with enhanced chemiluminescence substrate (ECL; Amersham Pharmacia Biotech). After exposure to the X-ray film, band intensities were quantified by densitometry and expressed as the mean area density using Scion Image (Scion Corporation, Frederick, MD). The mean area density is expressed relative to G3PDH (glyceraldehyde-3-phosphate dehydrogenase).
Measurement of DNIC by EPR spectrometry
To confirm the actual diffusion of RNOS from the mucosal side of the chamber into the mounted gastric tissue, the appearance of dinitrosyl dithiolato iron complex (DNIC) in the tissue was investigated using EPR spectroscopy at liquid nitrogen temperature. DNIC is thought to be generated by the reaction of RNOS (mainly NO) with iron within the tissue.41 The tissues were removed from the Ussing chamber after the experiments and were washed with saline, blotted, minced and then subjected to the measurement of DNIC by EPR spectroscopy as we recently reported.38 This sample was extruded into an EPR quartz tube (outer diameter, 5 mm) and immediately frozen in liquid nitrogen. To assess the intensity changes of the DNIC signal semi-quantitatively, the manganese (II), Mn (II), signal in the EPR spectra was used as an internal standard since Mn (II) signal intensities in the EPR spectra of animal tissues do not vary from sample to sample.38 X-band (∼9.2 GHz) EPR spectra were measured at liquid nitrogen temperature with a TE-200 EPR spectrometer (JEOL Ltd, Tokyo, Japan).
In vivo exposure
Anaesthetised rats were laid supine on a board and two fine polyethylene tubes (0.61 mm in diameter; Becton Dickinson, Franklin Lakes, NJ) were placed in the stomach via the mouth. One was used to administer pH 1.5 HCl aqueous solution (1.8 ml), and the other was used to administer sodium nitrite dissolved in saline (0.2 ml) to give a final concentration of 5.0 mM, or saline alone (0.2 ml) as a control. The procedure was repeated every 15 min for 2 h. After the luminal exposure in vivo, the stomach was removed and the gastric membrane was mounted in the Ussing chamber. Both sides of the chamber were bathed with modified Krebs buffer during the observation period, and the total transmembrane resistance and the [3H]mannitol flux were measured as described above.
Statistics
Data are expressed as the mean (SD) of the values in each group. Statistical analyses were performed using a one-way analysis of variance followed by the unpaired t test. Values associated with a probability (p value) of <0.05 were considered significant.
RESULTS
Transmembrane resistance
Changes in transmembrane resistance after administration of various concentrations of nitrite are shown in fig 2. In all of the nitrite-exposed mucosa, an initial rise in the transmembrane resistance was observed compared with the nitrite-free pH 2.5 acid controls. This initial rise in the transmembrane resistance may represent the increased resistance of the membrane to protect itself from apical insults, as reported previously.42 Thereafter, the transmembrane resistance was decreased rapidly when either 1.0 or 5.0 mM nitrite was administered, and the difference became significant compared with the acid controls at 60 min after the commencement of the nitrite administration. The administration of 0.1 mM nitrite also caused a gradual decrease in the transmembrane resistance, and the difference eventually became significant after 150 min.
The administration of pH 7.4 sodium nitrite or pH 2.5 sodium nitrate to the mucosal side of the chamber, both of which are known to be unsuitable conditions for RNOS formation, induced no detectable change in the transmembrane resistance during the observation period (data not shown). The addition of a reducing agent, ascorbic acid (10 mM), to the mucosal side of the chamber, coupled with nitrite plus pH 2.5 HCl, depleted the majority of the nitrite (approximately 90%) in the solution of the mucosal side within 1 min, resulting in no change in the gastric barrier function (data not shown).
Permeability of the gastric mucosa
Changes in [3H]mannitol flux from the mucosal to the serosal side after administration of various concentrations of nitrite are shown in fig 3. Compared with the nitrite-free acid controls, the [3H]mannitol flux was increased with time by the administration of nitrite. The increase in the flux became significant at 90 min for 1.0 or 5.0 mM nitrite, and at 120 min for 0.1 mM nitrite administration.
Histological findings
After the above experiments, the superficial epithelium of the mounted mucosa was found on routine microscopy to have remained intact in the 0.1 mM nitrite-exposed mucosa (fig 4B) compared with that in both the untreated controls (fig 4A) and the acid controls (fig 4B). However, in the gastric mucosa with the higher concentrations (1.0 and 5.0 mM) of nitrite exposure, morphological injury became evident in that the surface cells were deteriorated and detached from the membrane (fig 4D).
Immunostaining for occludin
Occludin was detected at apical cell adhesion sites, representing an intact tight junction as previously reported,43 and was observed in the nitrite-free controls (fig 5B) and in the untreated controls (fig 5A). Although administration of 0.1 mM nitrite did not cause histological damage to the gastric mucosa on routine microscopy, the expression of occludin in the superficial epithelium was fragmented or was only partially detected at that concentration of nitrite (fig. 5C). As expected from the microscopic findings that surface mucosal deterioration and detachment were evident in the mucosa with the 1.0 and 5.0 mM nitrite administration, the occludin expression had completely disappeared in such conditions (fig 5D).
Immunoblotting for occludin
The expression of occludin was evaluated by the relative mean area density of the band on western blotting. The occludin expression decreased significantly in the 0.1 mM nitrite-exposed mucosa compared with the untreated control and acid-treated control (p<0.05 for both). Although the acid treatment tended to decrease the occludin expression slightly, the difference was not significant (fig 6A,B).
Measurements of DNIC
The broad signal around g = 2.03 represents, judging from the characteristic line shape, a part of the DNIC spectrum as we reported recently.38 When the concentration of nitrite added was increased from 0.1 mM up to 5.0 mM with a fixed 3 h administration period, the signal intensity of DNIC in the gastric tissue was increased in a dose-dependent manner (fig 7A,B). These findings suggest that RNOS generated in the mucosal side of the chamber actually diffused into the mounted gastric membrane and reacted with iron within the tissue.
In vivo exposure
The total mucosal resistance of gastric membrane exposed in vivo to 5.0 mM nitrite at pH 1.5 HCl was significantly lowered throughout the observation period, compared with that of the nitrite-free pH 1.5 control (fig 8A). The [3H]mannitol flux from the mucosal to the serosal side of the chamber was increased with time more sharply in the gastric membrane with in vivo exposure to nitrite than in the nitrite-free control, and the difference became significant 60 min after commencing the measurement (fig 8B).
DISCUSSION
Using an ex vivo Ussing chamber technique, this study demonstrated that the gastric membrane whose mucosal side was exposed to a physiological concentration of acidified nitrite manifested a decrease in gastric electrical resistance with an increase in epithelial permeability, suggesting a disturbance of the gastric barrier function. Although these changes were observed even in the absence of detectable evidence of epithelial injury on routine microscopy, they were accompanied by a derangement of the expression of occludin, an integral membrane protein involved in epithelial tight junction formation. In addition, we also confirmed, by showing the generation of DNIC, a reaction product between RNOS and iron within the tissue, that RNOS induced by the acidification of nitrite administered to the mucosal side of the chamber diffused into the mounted gastric epithelium. Taken together, these observations suggest that RNOS generated in the acidic lumen diffuse into the adjacent tissue and derange the epithelial barrier function, at least partly by disrupting the tight junction.
Following the ingestion of a high nitrate meal, large amounts of nitrite are delivered into the human oesophagus and stomach by swallowing saliva.14–17 Since nitrite is stable in a neutral pH environment as in the non-refluxed oesophagus or the achlorhydric stomach, a high concentration of nitrite, a level similar to that seen in saliva, could be detected in such conditions.44–46 In contrast, it has been reported that there is little nitrite found in acid-secreting healthy stomach.44–46 Furthermore, a recent detailed study, which evaluated the regional concentrations of nitrite at various sites in the oesophagus and stomach, demonstrated that the nitrite concentration fell immediately on encountering gastric acid at the human GO junction,45 suggesting that once nitrite is acidified by gastric acid it is short-lived and is able to exist only at the GO junction. The disappearance of nitrite at the human GO junction may be explained by its rapid conversion into reactive RNOS, some of which are diffusible into the adjacent tissue.25 45 Accordingly, our present ex vivo model, in which physiological concentrations of nitrite were administered in pH 2.5 HCl solution bathing the mucosal side of the chamber, was designed to simulate the phenomenon occurring at the human GO junction.
Disruption of the gut barrier function by RNOS, in particular NO, has been documented to occur in previous in vitro and in vivo studies. Employing monolayers of the enterocytic cell line, Caco-2BBe, it has been shown that incubation of the monolayers with authentic NO gas or compounds that release NO markedly increases epithelial permeability.47 48 In in vivo studies using rats administered endotoxin, most of which investigated the intestinal barrier function, excess production of NO induced by iNOS was implicated in the loss of the gut barrier function.29–31 49 In addition, some ex vivo studies with an Ussing chamber technique have reported that NO administered as donor, whether placed serosally50 or mucosally,51 increased the intestinal membrane permeability. Employing the same chamber technique, the present study extended the findings of these previous studies by showing that the gastric membrane is considerably susceptible to mucosal exposure to RNOS generated by co-administration of nitrite plus pH 2.5 HCl although it is relatively resistant to acid alone.
Tight junctions are located at the luminal aspect of adjacent epithelial cells and form a barrier to the diffusion of solutes through the paracellular pathway.40 Since mannitol transverses the epithelium membrane only by way of a paracellular shunt, the increased permeability to mannitol, as observed in the present study, represents an alteration of the integrity of tight junctions. Occludin is one of the integral membrane proteins involved in tight junction formation, and depletion of the protein has been shown to correlate with both a decrease in transepithelial resistance and a perturbation of the permeability barrier.52 Recently, in a study that employed a Caco-2 monolayer, it was demonstrated that NO generated endogenously as the result of iNOS activity or exogenously from an NO donor decreased the expression of occludin, probably by a post-translational mechanism, concomitant with a derangement of the barrier function.53 Consistent with this, immunohistological and immunoblotting examination in the present study revealed that the expression of occludin was deranged in the gastric epithelium after a 3 h exposure to luminal acidified nitrite. Notably, exposure to 0.1 mM nitrite was sufficient to cause a derangement of occludin expression, concomitant with epithelial barrier dysfunction as demonstrated by both electrophysiological disturbance and an increase in epithelial permeability, although there was no obvious histological damage to the gastric mucosa at this concentration. Therefore, it is conceivable that luminal RNOS generated by acidification of nitrite impairs the adjacent gastric barrier function primarily by disrupting the tight junction. The finding that such a disruption of the gastric membrane could be initiated even at a concentration as low as 0.1 mM nitrite, a concentration observed even in some fasting humans,19 44 45 suggests that the human GO junction may be subjected to continuous chemical insult by rapid acidification of nitrite at that site not only after the ingestion of nitrate meal but all day long.
In contrast to the ex vivo exposure to RNOS employed in the chamber system of the present study, in the in vivo situation, the mucus layer, which lies between the acidic gastric lumen and the epithelial surface, could act as a pre-epithelial barrier and attenuate gastric barrier dysfunction by RNOS generated luminally. Accordingly, we tested whether in vivo exposure to luminal RNOS is sufficient to disturb the gastric barrier function using the same chamber technique. It was found that the exposure to luminal 5 mM nitrite in an acidic condition could cause a derangement of the gastric barrier function. Since this concentration of nitrite is still within the range observed in human saliva after the ingestion of a high nitrate meal, although the salivary nitrite is diluted to some extent when the saliva mixes with gastric acid,36–38 the results of the in vivo exposure experiment indicated that the gastric barrier dysfunction caused by RNOS could actually be induced in vivo, especially at the human GO junction.
Ascorbic acid, which is secreted in human gastric acid, reacts with nitrite in an acidic environment to form NO promptly.20 Because our present chamber system was open to air circulated with a gas-lift apparatus, the addition of ascorbic acid to the mucosal side of the chamber could effectively scavenge nitrite from the system by converting nitrite in the solution into the gas NO, and thereby successfully prevent RNOS-related gastric membrane damage. However, in the actual human GO junction, which is mostly closed to air, abundant NO generated luminally by the reaction of acidified nitrite with ascorbic acid could diffuse into the adjacent mucosa, leading to a more deleterious effect on the tissue.19 21
In this study, various kinds of RNOS were generated in the lumen by administering nitrite to the acid solution. However, which particular RNOS was actually entering the mucosa and damaging it is still conjectural. The causative species may be different depending on whether the exposure is ex vivo or in vivo. In the ex vivo exposure using the chamber system, the administered nitrite, upon acidification, will be present predominantly as nitrous acid and only a small percentage will be NO.20 Although NO is probably the most diffusible RNOS produced,24 it is unknown whether such a small amount of NO would be sufficient to account for the epithelial damage observed in the ex vivo exposure study. Alternatively, the nitrous acid abundantly present in the acidified lumen of the chamber may exert some effects on the gastric barrier function, independently of the NO-mediated mechanism. The experiment in which ascorbic acid was added to the acidified lumen of the chamber to convert nitrous acid to NO did not enable us to distinguish between the two possibilities, due to technical limitations of the current chamber system as mentioned above. In contrast, in the in vivo exposure study, the rat gastric mucosa secreted significant amounts of ascorbic acid,54 which causes the acidified nitrite and nitrous acid to be converted to NO at the epithelial luminal surface and thus potentially enter the mucosa as NO. Similarly, in human, since ascorbic acid is continuously secreted from the gastric mucosa,55 the acidified nitrite present in the lumen will eventually be converted to NO at the epithelial surface and diffuse into it, resulting in epithelial barrier dysfunction.
The gut barrier function is known to play a principal role in the maintenance of the mucosal structure and function in the presence of potentially damaging agents such as bile acid. Hence, disruption of the barrier function has been considered to be involved in the pathogenesis in a state of sepsis or inflammatory bowel disease.29–32 Similarly, persistent abnormalities in the barrier function at the human GO junction could be involved in the perpetuation of chronic inflammation at that site. The results of the present study suggested that disruption of the mucosal barrier function may represent the earliest effect of RNOS-derived stress on the epithelium at the GO junction, which is known to be exposed to abundant luminal RNOS. Considering the lifetime exposure of the GO junction to such cytotoxic levels of RNOS, this may be responsible for the high prevalence of inflammation and metaplasia, and subsequent development of neoplastic disease at that site.
In conclusion, simulating the microenvironment of the human GO junction, the results of this ex vivo chamber study suggest that RNOS generated luminally at the human GO junction can derange the gastric barrier function of the adjacent tissue by disrupting the tight junction. Thus, the tight junction of the surface epithelium at the GO junction may be a principal target of the RNOS arising from the lumen. The present chamber model will be useful for exploring the mechanism of RNOS-related cell injury occurring at the human GO junction by enabling direct evaluation of the interaction between the luminal chemistry and cell biology within the adjacent tissue.
REFERRENCE
Footnotes
Competing interests: None.
Funding: This work was supported in part by a Grant-in-Aid to TY for Scientific Research (17550146) from the Japan Society for the Promotion of Science and for Scientific Research on Priority Areas “Application of Molecular Spins” (15087212) from the Ministry of Education, Culture, Sports, Science and Technology, and in part by a Grant-in-Aid to KI (17790429) from the Ministry of Education, Science, Sports and Culture in Japan.