- Email: [email protected]
Calcitonin gene-related peptide affords gastric mucosal protection by activating potassium channel in Wistar rat
GASTROENTEROLOGY 1998;114:71–76
Calcitonin Gene-Related Peptide Affords Gastric Mucosal Protection by Activating Potassium Channel in Wistar Rat KOSEI DOI,* TETSUHIKO NAGAO,* KEISHI KAWAKUBO,* SETSURO IBAYASHI,* KUNIHIKO AOYAGI,* YUJI YANO,* CHIFUMI YAMAMOTO,* KOHKI KANAMOTO,* MITSUO IIDA,‡ SEIZO SADOSHIMA,* and MASATOSHI FUJISHIMA* *Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Fukuoka; and ‡Division of Gastroenterology, Department of Medicine, Kawasaki Medical School, Kurashiki, Japan
Background & Aims: Calcitonin gene-related peptide (CGRP) protects the gastric mucosa against injurious stimuli in various experimental models. The underlying mechanism could be the increase in gastric mucosal blood flow (GMBF). A number of endogenous vasodilators exert their effects through the activation of adenosine triphosphate (ATP)-sensitive potassium (KATP) channels on vascular smooth muscle. The present experiments were performed to elucidate whether CGRP increases GMBF through the activation of KATP channels and whether the channels are involved in the protection by CGRP of gastric mucosa. Methods: GMBF was determined by the hydrogen-clearance technique in male Wistar rats. Mucosal lesions were produced by intragastric superfusion with 0.15N HCl and 15% ethanol for 40 minutes. Effects of an agonist (Y-26763, intra-arterially) and an inhibitor (glibenclamide, intravenously) of KATP channels were tested. Results: Y-26763 increased GMBF, which was abolished by glibenclamide, and a CGRP-induced increase in GMBF was attenuated by glibenclamide. Macroscopic and microscopic lesions were exacerbated by human CGRP-(8– 37) (a CGRP-1 receptor antagonist; intra-arterially) and glibenclamide but were ameliorated by exogenous CGRP (intra-arterially). Conclusions: CGRP protects the gastric mucosa against ulcerogenic stimuli, at least in part, through the activation of KATP channels in rats.
he channel is an essential determinant factor of membrane potential and plays a major role in the regulation of arterial tone. Among divergent types of K1 channels, adenosine triphosphate (ATP)-sensitive potassium (KATP) channels are involved in the vasodilation in basilar, mesenteric, and coronary arteries.1–4 However, there are few reports regarding the role of KATP channels in the regulation of the gastric circulation. The first aim of this study was to examine whether KATP channels are functioning in arteries supplying the gastric mucosa. Postulated endogenous activators of KATP channels include calcitonin gene-related peptide (CGRP), nitric
T
oxide, and adenosine.5–7 Among these substances, CGRP mediates the increase in gastric mucosal blood flow (GMBF)8,9 and may play a major role in the defensive mechanisms of the gastric mucosa against pending acid injury.10 In rabbit mesenteric arteries, CGRP activates KATP channels2,5 and produces vasodilation, although the involvement of KATP channels in CGRP-induced vasodilation is not always the rule in other vascular beds.11 Our second goal was to examine whether CGRP uses KATP channels for its vasodilatory action in the gastric mucosa in rats. GMBF increases after the disruption of gastric mucosal barrier in the presence of acid.12–14 The increase in GMBF is explained, at least in part, by the release of CGRP from capsaicin-sensitive afferent neurons.10,15 Our final goal was to examine whether KATP channels are involved in the intrinsic protective mechanisms against the formation of mucosal lesion during intragastric acid challenge.
Materials and Methods Animal Preparation This experiment was reviewed by the Committee of the Ethics on Animal Experiment in the Faculty of Medicine, Kyushu University, and was performed under the control of the Guideline for Animal Experiment in the Faculty of Medicine, Kyushu University, and the law (no. 105) and notification (no. 6) of the Government. Male Wistar rats, weighing 240–330 g, were fasted for 24 hours but were allowed free access to water before experiments. After induction of anesthesia with amobarbital (100 mg/kg, intraperitoneally), rats underwent tracheostomy for spontaneous respiration and hydrogen gas inhalation. The femoral arteries on both sides were cannulated: one for the recording of arterial blood pressure and the other for the sampling of arterial blood. A femoral vein was cannulated on one side for the Abbreviations used in this paper: DMSO, dimethyl sulfoxide; GMBF, gastric mucosal blood flow; KATP, adenosine triphosphate– sensitive potassium; MABP, mean arterial blood pressure. r 1998 by the American Gastroenterological Association 0016-5085/98/$3.00
72
DOI ET AL.
continuous infusion of sorbitol (50 g/L) at a rate of 1.5 mL/h to avoid hypoglycemia caused by glibenclamide and dehydration. The rectal temperature was maintained at 37.07C 6 0.57C by a heating pad. In intra-arterial infusion experiments, a polyethylene catheter (PE-10) was inserted retrogradely into the splenic artery. The tip of the catheter was positioned close to the celiac artery. The stomach was cannulated by two polyethylene cannulas (OD, 2 mm), one for inflow and the other for outflow, and was perfused with warm HEPES buffer ([in mmol/L]: NaCl, 140; KCl, 4; CaCl2, 2.4; MgCl2, 2; HEPES, 10; and glucose, 11; pH 7.4 at 377C) at a rate of 45 mL/h.
Measurements of GMBF GMBF was measured by a hydrogen gas clearance method.16,17 After the pylorus was ligated, an incision was made in the forestomach, and the gastric contents were washed out gently with warm physiological salt solution (377C). A platinum sphere-type contact-electrode (1 3 1 mm; Unique Medical Co., Ltd., Tokyo, Japan) was fixed at the fundic posterior wall.18 Omeprazole was administered to inhibit possible stimulation of gastric acid secretion by glibenclamide, dimethyl sulfoxide (DMSO; a vehicle of glibenclamide), and Y-26763. Perfusate of the stomach was collected every 15 minutes. The pH of collected fluid was measured with a pH meter (Beckman f20; Beckman Instruments, Inc., Fullerton, CA), and the quantity of secreted proton was calculated. Gastric acid secretion was below a detectable level throughout the experiment.
Experiment 1 The experiments were performed to examine the effect of Y-26763, an opener of KATP channels.19 Fourteen rats were divided randomly into two groups: Y-26763 and glibenclamide 1 Y-26763 groups. Baseline GMBF was measured twice at 15-minute intervals. After measurements, glibenclamide (20 mg/kg, intravenously) or DMSO was administered. Two more GMBF measurements were made 5 and 15 minutes after the administration of glibenclamide or DMSO. Y-26763 (13.5 mmol/min) was infused for 5 minutes. GMBF was determined 5, 15, and 30 minutes after administration of Y-26763.
GASTROENTEROLOGY Vol. 114, No. 1
Experiment 3 Fifty-two rats were divided randomly into eight groups: CGRP, glibenclamide, CGRP-(8–37), glibenclamide 1 CGRP(8–37), and four respective vehicle control groups. CGRP (20 pmol/min) and CGRP-(8–37) (500 pmol/min, intra-arterially) was infused for 40 and 30 minutes, respectively. Glibenclamide (20 mg/kg) was administered into the femoral vein. After the treatments, the stomach was perfused with 0.15N HCl plus 15% ethanol for 40 minutes at a rate of 45 mL/h. After completing each experiment, the stomach was perfused with 1% formalin, cut along the large curvature of the stomach, and photographed. The photographs were taken into a Macintosh computer (LC 575). Mucosal lesions were quantified by means of National Institutes of Health Image (Ver. 1.56; National Institutes of Health, Bethesda, MD). We expressed the extent of lesion as follows: (Area of Lesion/Glandular Area) 3 100 (%).18 The tissues were then processed for histological analysis. After fixation for 5–7 days, a standardized strip of tissue across the anterior wall of the proximal corpus was removed, embedded, sectioned, stained, and evaluated microscopically. Three grades of histological injury were recorded20: no damage, shallow damage involving up to 25% of the mucosal depth, and deep damage involving .25% of the mucosal depth. The histological injury grades were expressed as percent length of the section occupied by lesions with respective injury grades.
Drugs Y-26763 (Yoshitomi Pharmaceuticals Ind., Ltd., Fukuoka, Japan) has pharmacological properties similar to those of levcromakalim, a prototype opener of KATP channels, and is regarded as one of the selective openers of the channels.19 Glibenclamide (Sigma Chemical Co., St. Louis, MO) was dissolved in DMSO (final concentration, ,0.01% volume) and was infused into the femoral vein as a bolus injection. Rat CGRP and CGRP-(8–37) were obtained from Sigma Chemical Co. and were dissolved in distilled water. Y-26763, CGRP, and CGRP-(8–37) were infused into the celiac artery at the rate of 25 mL/min. Omeprazole (Fujisawa Pharmaceutical Co., Ltd., Osaka, Japan) was dissolved in 0.5% methyl cellulose and 0.2% NaHCO3-NaOH solution.
Experiment 2 Experiments were designed to test the hypothesis that CGRP activates KATP channels. Forty rats were divided into six groups in a random fashion: CGRP, glibenclamide 1 CGRP, CGRP-(8–37) (an antagonist of CGRP-1 receptors) 1 CGRP groups, and three respective vehicle control groups. Glibenclamide (20 mg/kg) was administered into the femoral vein as a bolus injection after two baseline GMBF measurements. CGRP-(8–37) (500 pmol/min) was infused for 30 minutes after baseline GMBF measurements. Two more GMBF measurements were made after glibenclamide or during CGRP-(8–37). Then, CGRP (20 pmol/min) was infused into the celiac artery for 10 minutes. GMBF was determined at 5 and 15 minutes after the introduction of CGRP. Three respective vehicle control experiments were performed.
Statistics The data were expressed as means 6 SD. Statistical differences both among and within groups were analyzed with a two-factor repeated measures analysis of variance followed by Fisher’s protected least significant difference test. P values of ,0.05 were regarded as statistically significant.
Results Experiment 1 Physiological variables. There were no differ-
ences in physiological variables at baseline and pretreatment periods of Y-26763 between the two groups of rats. Arterial blood gas parameters and hematocrit were at
January 1998
ATP–SENSITIVE POTASSIUM CHANNEL AND CGRP
73
constant levels throughout the experiment. The blood glucose level was lower in the glibenclamide 1 Y-26763 group but was within a physiological limit (Table 1). Responses to Y-26763. In both groups, mean arterial blood pressure (MABP) was within a physiological variation throughout the experiment (Figure 1A). Baseline GMBF was 80.5 6 17.2 and 81.7 6 19.1 mL · min21· 100 g tissue21 in the Y-26763 group and glibenclamide 1 Y-26763 group, respectively. GMBF did not change significantly after administration of glibenclamide or its vehicle. GMBF increased by 50% 6 8.4% at 5 minutes after intra-arterial administration of Y-26763 (13.5 mmol/min; Figure 1B). The increase in GMBF induced by Y-26763 was abolished by glibenclamide (20 mg/kg) but not by DMSO (data not shown). Experiment 2 Physiological variables. There were no differ-
ences in baseline physiological variables among the six groups of rats. Although the blood glucose level in the glibenclamide 1 CGRP group was lower than that in the other groups, the parameter was within a physiological range (Table 2). Role of KATP channels in CGRP-induced GMBF increase. Baseline GMBF was 78.6 6 25.1, 75.1 6
25.7, and 71.6 6 21.8 mL · min 21 · 100 g tissue21 in the CGRP, glibenclamide 1 CGRP, and CGRP-(8–37) 1 CGRP groups, respectively. GMBF increased to 110 6 20.4 mL · min21 · 100 g tissue21 after administration of CGRP (20 pmol/min; intra-arterially). The increase in Table 1. Physiological Variables in Experiment 1 Y-26763 Baseline
Pretreatment
After 5 min
PaCO2 (mm Hg) Y 36.6 6 0.8 35.9 6 1.1 35.1 6 3.6 a G1Y 38.4 6 2.0 37.1 6 2.7 33.3 6 1.2 a PaO2 (mm Hg) Y 83.4 6 5.7 84.0 6 3.1 85.5 6 6.8 G1Y 82.5 6 7.1 81.8 6 6.6 85.5 6 4.9 pH Y 7.37 6 0.011 7.39 6 0.008 a 7.41 6 0.008a,b G1Y 7.38 6 0.024 7.39 6 0.027 7.41 6 0.020 a Hematocrit (% ) Y 43.2 6 1.4 43.9 6 1.1 42.8 6 1.1 G1Y 43.3 6 2.4 43.8 6 1.7 43.7 6 2.0 Glucose (mg/dL) Y 107.4 6 11.6 107.0 6 13.9 114.9 6 13.0 G1Y 96.3 6 19.6 84.5 6 6.9 c 85.3 6 8.8 c NOTE. Values are expressed as means 6 SD. G, glibenclamide (20 mg/kg, intravenously); Y, Y-26763 (13.5 mmol/ min, intra-arterially). aP , 0.05 vs. baseline. bP , 0.05 vs. pretreatment. cP , 0.001 vs. Y-26763.
Figure 1. Changes in (A) MABP and (B) GMBF in Y-26763 group (X; n 5 8) and glibenclamide 1 Y-26763 group (N; n 5 6) in experiment 1. *P , 0.05 vs. glibenclamide 1 Y-26763; **P , 0.05 vs. baseline and pretreatment in the same group. Bars represent SD. ia, intraarterially; iv, intravenously.
GMBF by CGRP was 45% of the pretreatment values (Figure 2B), despite a decrease in MABP by 12 mm Hg (Figure 2A). In the glibenclamide-treated group, GMBF increased by 16% during administration of CGRP (Figure 2B). The effect of CGRP on GMBF and MABP was almost absent in the CGRP-(8–37)–treated (500 pmol/min) group. In three respective vehicle control groups, MABP and GMBF were stable throughout the experiments (data not shown). Experiment 3 Effect of glibenclamide on gastric mucosal injury. Macroscopic examination. Intragastric perfusion with
0.15N HCl 1 15% ethanol induced hemorrhagic lesions in the glandular area of the stomach. The lesions in the glibenclamide group were larger than those in the vehicle control (DMSO) group but were smaller than those in the CGRP-(8–37) group (Figure 3).21 The lesions in the CGRP-(8–37) group and the glibenclamide 1 CGRP-(8– 37) group were comparable but were larger than those in the corresponding control groups (Figure 3). The lesions in the CGRP group were smaller than those in the corresponding vehicle control group (Figure 3). Microscopic examination. The proportion of section length occupied by deep damage was in the following ascending order: CGRP, control, glibenclamide, and CGRP-(8–37) groups (Figure 4). The percentages of
74
DOI ET AL.
GASTROENTEROLOGY Vol. 114, No. 1
Table 2. Physiological Variables in Experiment 2 CGRP Baseline PaCO2 (mm Hg) C 35.7 6 2.3 G1C 36.2 6 1.4 C(8–37) 1 C 34.6 6 2.0 PaO2 (mm Hg) C 80.5 6 5.7 G1C 81.8 6 8.2 C(8–37) 1 C 80.6 6 6.9 pH C 7.37 6 0.011 G1C 7.38 6 0.028 C(8–37) 1 C 7.36 6 0.029 Hematocrit (% ) C 44.5 6 1.4 G1C 42.8 6 2.8 C(8–37) 1 C 43.5 6 2.7 Glucose (mg/dL) C 100.4 6 11.6 G1C 104.6 6 22.6 C(8–37) 1 C 98.5 6 9.6
Pretreatment
After 5 min
36.2 6 1.7 36.5 6 1.7 33.1 6 2.5
34.4 6 3.7 35.1 6 4.2 32.2 6 2.2
83.5 6 1.4 83.5 6 1.4 81.2 6 4.7
84.2 6 5.4 a 84.2 6 5.4 a 84.1 6 7.1 a,b
7.39 6 0.096 7.39 6 0.031 7.38 6 0.24
7.43 6 0.051 a,b 7.43 6 0.104 a 7.42 6 0.49 a
45.7 6 1.7 44.0 6 4.8 44.5 6 3.7
44.2 6 3.4 43.0 6 3.7 42.9 6 4.9
98.0 6 13.9 90.1 6 7.9 c 97.9 6 11.5
108.1 6 13.0 88.9 6 23.8 a,c 100.8 6 16.9
NOTE. Values are means 6 SD. C, CGRP (20 pmol/min, intra-arterially); G, glibenclamide (20 mg/kg, intravenously); C(8–37), CGRP-(8–37) (500 pmol/min, intra-arterially). aP , 0.05 vs. baseline. bP , 0.05 vs. pretreatment. cP , 0.001 vs. CGRP and CGRP (8–37) 1 C.
lesion with shallow damage in CGRP, control, glibenclamide, and CGRP-(8–37) groups were 55.4% 6 11.1%, 40.5% 6 12.5%, 38.9% 6 19.2%, and 30.9% 6 24.3%, whereas those without detectable damage were 34.6% 6 6.6%, 30.1% 6 10.2%, 22.0% 6 15.6%, and 20.9% 6 25.3%, respectively.
Discussion There are three major findings in the present study. First, the activation of KATP channels increases GMBF. Second, a part of the increase in GMBF by CGRP is mediated by the activation of KATP channels. Finally, endogenous CGRP may protect the gastric mucosa against acid challenge, at least in part, through the activation of KATP channels. In some previous studies in which GMBF was measured by a hydrogen clearance method, GMBF was approximately 40 mL · min21 · 100 g tissue21 in rats.8,22 The value is lower than that reported in the present study. The inconsistency most likely results from the methodological difference. We used a contact-type electrode, whereas an insertion-type electrode was used in the previous studies. The value reported in the present study is similar to GMBF estimated by a contact-type electrode.17
Figure 2. Changes in (A) MABP and (B) GMBF in CGRP group (X; n 5 8), glibenclamide 1 CGRP group (N; n 5 8), and CGRP-(8–37) 1 CGRP group (S; n 5 6) in experiment 2. *P , 0.05 vs. glibenclamide 1 CGRP; †P , 0.05 vs. baseline and pretreatment in the same group; **P , 0.05 vs. CGRP-(8–37) 1 CGRP. Bars represent SD. ia, intra-arterially; iv, intravenously.
Intragastric perfusion with 0.15N HCl plus 15% ethanol produces gastric mucosal lesion histologically similar to human peptic ulcer. Furthermore, an increase in acid backdiffusion, which is known as one of the underlying mechanisms for peptic ulcer, is also reported in this model.14 Some investigations have been performed using this model to study mechanisms for gastric mucosal protection.10,14 These facts should validate the use of this model in the present experiments. Y-26763, CGRP, and CGRP-(8–37) were administered by a continuous infusion into the celiac artery. This procedure would be the best way to clarify the effect of these agents on the gastric mucosal circulation because any change in systemic hemodynamics such as blood pressure can be minimized with an intra-arterial injection. By contrast, glibenclamide was administered into the femoral vein as a bolus injection. This is based on previous reports that showed potent and long-lasting blockade of KATP channels with this procedure.23,24 Activation of KATP channels is an important mechanism for vasodilation.2,5 There are few studies, however, about the role of KATP channels in gastric circulation.25 In the present study, intra-arterial infusion of Y-26763 increased GMBF without changing systemic blood pres-
January 1998
Figure 3. Macroscopic gastric mucosal injury. Lesion Index (%) 5 (Lesion Area/Total Glandular Area) 3 100. *P , 0.05. Bars represent means 6 SD.
sure, and the responses were inhibited by glibenclamide, a selective inhibitor of KATP channels.26,27 These results suggest that KATP channels are present and functioning in gastric mucosal circulation. KATP channels play a role in the determination of resting tone in some3,4,28 but not all arteries or vascular beds.29 The present study showed that baseline GMBF was not affected by glibenclamide alone. This observation suggests that KATP channels do not contribute to the determination of baseline GMBF in Wistar rats. However, we should take the effect of anesthesia and surgical procedures into account. The involvement of KATP channels in vasodilation induced by CGRP is heterogeneous among different
Figure 4. Microscopic gastric mucosal injury. The vertical line represents the proportion of deep damage in examined sections. Deep damage refers to mucosal lesions affecting deeper than 25% of the mucosa. *P , 0.05. Bars represent means 6 SD. hCGRP-(8–37), human CGRR-(8–37).
ATP–SENSITIVE POTASSIUM CHANNEL AND CGRP
75
vascular beds.1,2,28 The increase in GMBF induced by CGRP, at the dose used in the present study, was inhibited by 64% in the presence of glibenclamide in rats. The result is consistent with the view that CGRP increases GMBF, at least in part, by activating KATP channels in the gastric vascular bed of rats. Because the increase in GMBF to CGRP was abolished by CGRP-(8– 37), the opening of KATP channels should be mediated by the activation of CGRP-1 receptors. Two major mechanisms by which CGRP activates KATP channels have been postulated: activation of adenylate cyclase in vascular smooth muscle2 and release of NO from the endothelium.30 It is possible that endothelial NO activates KATP channels on arterial smooth muscle, as shown in mesenteric arteries of rabbits.6 CGRP-induced increase in GMBF is believed to be an important defense mechanism of the gastric mucosa against ulcerogenic factors.10 Thus, KATP channels may also be involved in mucosal defense as an antiulcerogenic factor. In fact, glibenclamide and CGRP-(8–37) worsened gastric mucosal injuries in the present study. Furthermore, glibenclamide had no additional effects on CGRP-(8–37). This observation is consistent with the view that KATP channels are activated by CGRP and protect gastric mucosa during acid challenge. Thus, KATP channel openers may have a protective effect on gastric mucosa by increasing GMBF. The present study warrants further investigations for the clinical usefulness of KATP channel openers as an antiulcer drug.
References 1. Kitazono T, Heisted D, Faraci FM. Role of ATP-sensitive K1 channels in CGRP-induced dilation of basilar artery in vivo. Am J Physiol 1993;265:H581–H585. 2. Quayle MJ, Bonev AD, Brayden JE, Nelson MT. Calcitonin generelated peptide activated ATP-sensitive K1 currents in rabbit arterial smooth muscle via protein kinase A. J Physiol Lond 1994;475:9–13. 3. Nagao T, Sadoshima S, Kamouchi M, Fujishima M. Cromakalim dilates rat cerebral arteries in vivo. Stroke 1991;22:221–224. 4. Imamura Y, Tomoike H, Narishige T, Takahashi T, Kasuya H, Takeshita T. Glibenclamide decreases basal coronary blood flow in anesthetized dogs. Am J Physiol 1992;263:H399–H404. 5. Nelson MT, Huang Y, Brayden JE, Hescheler J, Standen NB. Arterial dilations in response to calcitonin gene-related peptide involve activation of K1 channels. Nature 1990;344:770–773. 6. Murphy EM, Brayden JE. Nitric oxide hyperpolarized rabbit mesenteric arteries via ATP-sensitive potassium channels. J Physiol Lond 1995;486:47–58. 7. Dart C, Standen NB. Adenosine-activated potassium current in smooth muscle cells isolated from the pig coronary artery. J Physiol Lond 1993;471:767–786. 8. Li DS, Raybould HE, Quintero E, Guth PH. Role of calcitonin gene-related peptide in gastric hyperemic response to intragastric capsaicin. Am J Physiol 1991;261:G657–G661. 9. Holzer P, Guth PH. Neuropeptide control of rat gastric mucosal blood flow: increase by calcitonin gene-related peptide and
76
10.
11.
12. 13. 14.
15.
16. 17.
18.
19.
20.
21.
DOI ET AL.
vasoactive intestinal peptide, but not substance P and neurokinin A. Circ Res 1991;68:100–105. Li DS, Raybould HE, Quintero E, Guth PH. Calcitonin gene-related peptide mediates the gastric hyperemic response to acid backdiffusion. Gastroenterology 1992;102:1124–1128. Zschauer A, Uusitalo H, Brayden JE. Role of endothelium and hyperpolarization in CGRP-induced vasodilation of rabbit ophthalmic artery. Am J Physiol 1992;263:H359–H365. Davenport HW. The gastric mucosal barrier. Digestion 1972;5: 162–165. Ritchie WP. Acute gastric mucosal damage induced by bile salts, acid and ischemia. Gastroenterology 1975;68:699–707. Holzer P, Livingston EH, Guth PH. Sensory neurons signal for an increase in rat gastric mucosal blood flow in the face of pending acid injury. Gastroenterology 1991;101:416–423. Chen R, Li DS, Guth PH. Capsaicin-sensitive afferent nerve induced gastric arteriolar dilatation is mediated in rat by calcitonin gene-related peptide (CGRP) (abstr). Gastroenterology 1991; 100:A42. Aukland K, Bower BF, Berliner RW. Measurement of local blood flow with hydrogen gas. Circ Res 1964;14:164–187. Murakami M, Morinaga M, Mitake T, Uchino H. Contact electrode method in hydrogen gas clearance technique: a new method for determination of regional gastric mucosal blood flow in animals and humans. Gastroenterology 1982;82:457–467. Kawakubo K, Ibayashi S, Nagao T, Doi K, Aoyagi K, Iida M, Sadoshima S, Fujishima M. Brain ischemia decreases gastric mucosal blood flow in hypertensive rats: role of arterial vagal adrenoceptors. Dig Dis Sci 1996;41:2383–2391. Nakajima T, Shinohara T, Yaoka O, Fukunari A, Shinagawa K, Aoki K, Katoh A, Yamanaka T, Setoguchi M, Tahara T. Y-27152, a long-acting K1 channel opener with less tachycardia: antihypertensive effects in hypertensive rats and dogs in conscious state. J Pharmacol Exp Ther 1992;261:730–736. Holzer P, Pabst MA, Lippe IT, Peskar BM, Peskar BA, Livingston EH, Guth PH. Afferent nerve-mediated protection against deep mucosal damage in the rat stomach. Gastroenterology 1990;98: 838–848. Doi K, Nagao T, Kawakubo K, Yano Y, Yamamoto C, Kanamoto K, Aoyagi K, Ibayashi T, Fujishima M. Calcitonin gene-related peptide protects against the formation of gastric mucosal lesions through the activation of ATP-sensitive potassium channel (abstr). Gastroenterology 1996;110:A96.
GASTROENTEROLOGY Vol. 114, No. 1
22. Chen YZR, Guth PH. Interaction of endogenous nitric oxide and CGRP in sensory neuron-induced gastric vasodilation. Am J Physiol 1995;268:791–796. 23. Thornton JD, Thornton CS, Sterling DL, Downey JM. Blockade of ATP-sensitive potassium channels increases infarct size but does not prevent preconditioning in rabbit hearts. Circ Res 1993;72: 44–49. 24. Furukawa S, Satoh K, Taira N. Opening of ATP-sensitive K1 channels responsible for adenosine A2 receptor–mediated vasodepression does not involve a pertussis toxin–sensitive G protein. Eur J Pharmacol 1993;236:255–262. 25. Cook NS, Fozard JR, Hof RP. The potential of potassium channel openers in peripheral vascular disease. In: Escande D, Standen N, eds. K1 channels in cardiovascular medicine. Paris: SpringerVerlag, 1993:305–319. 26. Ashford MLJ, Sturgess NC, Trout NJ, Gardner NJ, Hales CN. Adenosine 58-triphosphate–sensitive ion channels in neonatal rat culture central neurons. Pflugers Arch 1988;412:297–304. 27. Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT. Hyperpolarizing vasodilators activate ATP-sensitive K1 channels in arterial smooth muscle. Science 1989;245:177–180. 28. Nagao T, Ibayashi S, Sadoshima S, Fujii K, Fujii K, Ohya Y, Fujishima M. Distribution and physiological roles of ATP-sensitive K1 channels in the vertebrobasilar system of the rabbit. Circ Res 1996;78:238–243. 29. McPherson GA, Stork AP. The resistance of some rat cerebral arteries to the vasorelaxant effect of cromakalim and other K1 channel openers. Br J Pharmacol 1992;105:51–58. 30. Whittle BJR, Lopez-Belmonte J, Moncada S. Nitric oxide mediates rat mucosal vasodilatation induced by intragastric capsaicin. Eur J Pharmacol 1992;218:339–341.
Received October 30, 1996. Accepted September 16, 1997. Address requests for reprints to: Kosei Doi, M.D., Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka 812-82, Japan. Fax: (81) 92-642-5273. Presented in part at Digestive Disease Week in San Francisco, California, on May 19–22, 1996. The authors thank Yoshitomi Pharmaceuticals Ind., Ltd., Fukuoka, Japan, for providing Y-26763; and Fujisawa Pharmaceutical Co. Ltd., Osaka, Japan, for providing omeprazole.
Calcitonin Gene-Related Peptide Affords Gastric Mucosal Protection by Activating Potassium Channel in Wistar Rat KOSEI DOI,* TETSUHIKO NAGAO,* KEISHI KAWAKUBO,* SETSURO IBAYASHI,* KUNIHIKO AOYAGI,* YUJI YANO,* CHIFUMI YAMAMOTO,* KOHKI KANAMOTO,* MITSUO IIDA,‡ SEIZO SADOSHIMA,* and MASATOSHI FUJISHIMA* *Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Fukuoka; and ‡Division of Gastroenterology, Department of Medicine, Kawasaki Medical School, Kurashiki, Japan
Background & Aims: Calcitonin gene-related peptide (CGRP) protects the gastric mucosa against injurious stimuli in various experimental models. The underlying mechanism could be the increase in gastric mucosal blood flow (GMBF). A number of endogenous vasodilators exert their effects through the activation of adenosine triphosphate (ATP)-sensitive potassium (KATP) channels on vascular smooth muscle. The present experiments were performed to elucidate whether CGRP increases GMBF through the activation of KATP channels and whether the channels are involved in the protection by CGRP of gastric mucosa. Methods: GMBF was determined by the hydrogen-clearance technique in male Wistar rats. Mucosal lesions were produced by intragastric superfusion with 0.15N HCl and 15% ethanol for 40 minutes. Effects of an agonist (Y-26763, intra-arterially) and an inhibitor (glibenclamide, intravenously) of KATP channels were tested. Results: Y-26763 increased GMBF, which was abolished by glibenclamide, and a CGRP-induced increase in GMBF was attenuated by glibenclamide. Macroscopic and microscopic lesions were exacerbated by human CGRP-(8– 37) (a CGRP-1 receptor antagonist; intra-arterially) and glibenclamide but were ameliorated by exogenous CGRP (intra-arterially). Conclusions: CGRP protects the gastric mucosa against ulcerogenic stimuli, at least in part, through the activation of KATP channels in rats.
he channel is an essential determinant factor of membrane potential and plays a major role in the regulation of arterial tone. Among divergent types of K1 channels, adenosine triphosphate (ATP)-sensitive potassium (KATP) channels are involved in the vasodilation in basilar, mesenteric, and coronary arteries.1–4 However, there are few reports regarding the role of KATP channels in the regulation of the gastric circulation. The first aim of this study was to examine whether KATP channels are functioning in arteries supplying the gastric mucosa. Postulated endogenous activators of KATP channels include calcitonin gene-related peptide (CGRP), nitric
T
oxide, and adenosine.5–7 Among these substances, CGRP mediates the increase in gastric mucosal blood flow (GMBF)8,9 and may play a major role in the defensive mechanisms of the gastric mucosa against pending acid injury.10 In rabbit mesenteric arteries, CGRP activates KATP channels2,5 and produces vasodilation, although the involvement of KATP channels in CGRP-induced vasodilation is not always the rule in other vascular beds.11 Our second goal was to examine whether CGRP uses KATP channels for its vasodilatory action in the gastric mucosa in rats. GMBF increases after the disruption of gastric mucosal barrier in the presence of acid.12–14 The increase in GMBF is explained, at least in part, by the release of CGRP from capsaicin-sensitive afferent neurons.10,15 Our final goal was to examine whether KATP channels are involved in the intrinsic protective mechanisms against the formation of mucosal lesion during intragastric acid challenge.
Materials and Methods Animal Preparation This experiment was reviewed by the Committee of the Ethics on Animal Experiment in the Faculty of Medicine, Kyushu University, and was performed under the control of the Guideline for Animal Experiment in the Faculty of Medicine, Kyushu University, and the law (no. 105) and notification (no. 6) of the Government. Male Wistar rats, weighing 240–330 g, were fasted for 24 hours but were allowed free access to water before experiments. After induction of anesthesia with amobarbital (100 mg/kg, intraperitoneally), rats underwent tracheostomy for spontaneous respiration and hydrogen gas inhalation. The femoral arteries on both sides were cannulated: one for the recording of arterial blood pressure and the other for the sampling of arterial blood. A femoral vein was cannulated on one side for the Abbreviations used in this paper: DMSO, dimethyl sulfoxide; GMBF, gastric mucosal blood flow; KATP, adenosine triphosphate– sensitive potassium; MABP, mean arterial blood pressure. r 1998 by the American Gastroenterological Association 0016-5085/98/$3.00
72
DOI ET AL.
continuous infusion of sorbitol (50 g/L) at a rate of 1.5 mL/h to avoid hypoglycemia caused by glibenclamide and dehydration. The rectal temperature was maintained at 37.07C 6 0.57C by a heating pad. In intra-arterial infusion experiments, a polyethylene catheter (PE-10) was inserted retrogradely into the splenic artery. The tip of the catheter was positioned close to the celiac artery. The stomach was cannulated by two polyethylene cannulas (OD, 2 mm), one for inflow and the other for outflow, and was perfused with warm HEPES buffer ([in mmol/L]: NaCl, 140; KCl, 4; CaCl2, 2.4; MgCl2, 2; HEPES, 10; and glucose, 11; pH 7.4 at 377C) at a rate of 45 mL/h.
Measurements of GMBF GMBF was measured by a hydrogen gas clearance method.16,17 After the pylorus was ligated, an incision was made in the forestomach, and the gastric contents were washed out gently with warm physiological salt solution (377C). A platinum sphere-type contact-electrode (1 3 1 mm; Unique Medical Co., Ltd., Tokyo, Japan) was fixed at the fundic posterior wall.18 Omeprazole was administered to inhibit possible stimulation of gastric acid secretion by glibenclamide, dimethyl sulfoxide (DMSO; a vehicle of glibenclamide), and Y-26763. Perfusate of the stomach was collected every 15 minutes. The pH of collected fluid was measured with a pH meter (Beckman f20; Beckman Instruments, Inc., Fullerton, CA), and the quantity of secreted proton was calculated. Gastric acid secretion was below a detectable level throughout the experiment.
Experiment 1 The experiments were performed to examine the effect of Y-26763, an opener of KATP channels.19 Fourteen rats were divided randomly into two groups: Y-26763 and glibenclamide 1 Y-26763 groups. Baseline GMBF was measured twice at 15-minute intervals. After measurements, glibenclamide (20 mg/kg, intravenously) or DMSO was administered. Two more GMBF measurements were made 5 and 15 minutes after the administration of glibenclamide or DMSO. Y-26763 (13.5 mmol/min) was infused for 5 minutes. GMBF was determined 5, 15, and 30 minutes after administration of Y-26763.
GASTROENTEROLOGY Vol. 114, No. 1
Experiment 3 Fifty-two rats were divided randomly into eight groups: CGRP, glibenclamide, CGRP-(8–37), glibenclamide 1 CGRP(8–37), and four respective vehicle control groups. CGRP (20 pmol/min) and CGRP-(8–37) (500 pmol/min, intra-arterially) was infused for 40 and 30 minutes, respectively. Glibenclamide (20 mg/kg) was administered into the femoral vein. After the treatments, the stomach was perfused with 0.15N HCl plus 15% ethanol for 40 minutes at a rate of 45 mL/h. After completing each experiment, the stomach was perfused with 1% formalin, cut along the large curvature of the stomach, and photographed. The photographs were taken into a Macintosh computer (LC 575). Mucosal lesions were quantified by means of National Institutes of Health Image (Ver. 1.56; National Institutes of Health, Bethesda, MD). We expressed the extent of lesion as follows: (Area of Lesion/Glandular Area) 3 100 (%).18 The tissues were then processed for histological analysis. After fixation for 5–7 days, a standardized strip of tissue across the anterior wall of the proximal corpus was removed, embedded, sectioned, stained, and evaluated microscopically. Three grades of histological injury were recorded20: no damage, shallow damage involving up to 25% of the mucosal depth, and deep damage involving .25% of the mucosal depth. The histological injury grades were expressed as percent length of the section occupied by lesions with respective injury grades.
Drugs Y-26763 (Yoshitomi Pharmaceuticals Ind., Ltd., Fukuoka, Japan) has pharmacological properties similar to those of levcromakalim, a prototype opener of KATP channels, and is regarded as one of the selective openers of the channels.19 Glibenclamide (Sigma Chemical Co., St. Louis, MO) was dissolved in DMSO (final concentration, ,0.01% volume) and was infused into the femoral vein as a bolus injection. Rat CGRP and CGRP-(8–37) were obtained from Sigma Chemical Co. and were dissolved in distilled water. Y-26763, CGRP, and CGRP-(8–37) were infused into the celiac artery at the rate of 25 mL/min. Omeprazole (Fujisawa Pharmaceutical Co., Ltd., Osaka, Japan) was dissolved in 0.5% methyl cellulose and 0.2% NaHCO3-NaOH solution.
Experiment 2 Experiments were designed to test the hypothesis that CGRP activates KATP channels. Forty rats were divided into six groups in a random fashion: CGRP, glibenclamide 1 CGRP, CGRP-(8–37) (an antagonist of CGRP-1 receptors) 1 CGRP groups, and three respective vehicle control groups. Glibenclamide (20 mg/kg) was administered into the femoral vein as a bolus injection after two baseline GMBF measurements. CGRP-(8–37) (500 pmol/min) was infused for 30 minutes after baseline GMBF measurements. Two more GMBF measurements were made after glibenclamide or during CGRP-(8–37). Then, CGRP (20 pmol/min) was infused into the celiac artery for 10 minutes. GMBF was determined at 5 and 15 minutes after the introduction of CGRP. Three respective vehicle control experiments were performed.
Statistics The data were expressed as means 6 SD. Statistical differences both among and within groups were analyzed with a two-factor repeated measures analysis of variance followed by Fisher’s protected least significant difference test. P values of ,0.05 were regarded as statistically significant.
Results Experiment 1 Physiological variables. There were no differ-
ences in physiological variables at baseline and pretreatment periods of Y-26763 between the two groups of rats. Arterial blood gas parameters and hematocrit were at
January 1998
ATP–SENSITIVE POTASSIUM CHANNEL AND CGRP
73
constant levels throughout the experiment. The blood glucose level was lower in the glibenclamide 1 Y-26763 group but was within a physiological limit (Table 1). Responses to Y-26763. In both groups, mean arterial blood pressure (MABP) was within a physiological variation throughout the experiment (Figure 1A). Baseline GMBF was 80.5 6 17.2 and 81.7 6 19.1 mL · min21· 100 g tissue21 in the Y-26763 group and glibenclamide 1 Y-26763 group, respectively. GMBF did not change significantly after administration of glibenclamide or its vehicle. GMBF increased by 50% 6 8.4% at 5 minutes after intra-arterial administration of Y-26763 (13.5 mmol/min; Figure 1B). The increase in GMBF induced by Y-26763 was abolished by glibenclamide (20 mg/kg) but not by DMSO (data not shown). Experiment 2 Physiological variables. There were no differ-
ences in baseline physiological variables among the six groups of rats. Although the blood glucose level in the glibenclamide 1 CGRP group was lower than that in the other groups, the parameter was within a physiological range (Table 2). Role of KATP channels in CGRP-induced GMBF increase. Baseline GMBF was 78.6 6 25.1, 75.1 6
25.7, and 71.6 6 21.8 mL · min 21 · 100 g tissue21 in the CGRP, glibenclamide 1 CGRP, and CGRP-(8–37) 1 CGRP groups, respectively. GMBF increased to 110 6 20.4 mL · min21 · 100 g tissue21 after administration of CGRP (20 pmol/min; intra-arterially). The increase in Table 1. Physiological Variables in Experiment 1 Y-26763 Baseline
Pretreatment
After 5 min
PaCO2 (mm Hg) Y 36.6 6 0.8 35.9 6 1.1 35.1 6 3.6 a G1Y 38.4 6 2.0 37.1 6 2.7 33.3 6 1.2 a PaO2 (mm Hg) Y 83.4 6 5.7 84.0 6 3.1 85.5 6 6.8 G1Y 82.5 6 7.1 81.8 6 6.6 85.5 6 4.9 pH Y 7.37 6 0.011 7.39 6 0.008 a 7.41 6 0.008a,b G1Y 7.38 6 0.024 7.39 6 0.027 7.41 6 0.020 a Hematocrit (% ) Y 43.2 6 1.4 43.9 6 1.1 42.8 6 1.1 G1Y 43.3 6 2.4 43.8 6 1.7 43.7 6 2.0 Glucose (mg/dL) Y 107.4 6 11.6 107.0 6 13.9 114.9 6 13.0 G1Y 96.3 6 19.6 84.5 6 6.9 c 85.3 6 8.8 c NOTE. Values are expressed as means 6 SD. G, glibenclamide (20 mg/kg, intravenously); Y, Y-26763 (13.5 mmol/ min, intra-arterially). aP , 0.05 vs. baseline. bP , 0.05 vs. pretreatment. cP , 0.001 vs. Y-26763.
Figure 1. Changes in (A) MABP and (B) GMBF in Y-26763 group (X; n 5 8) and glibenclamide 1 Y-26763 group (N; n 5 6) in experiment 1. *P , 0.05 vs. glibenclamide 1 Y-26763; **P , 0.05 vs. baseline and pretreatment in the same group. Bars represent SD. ia, intraarterially; iv, intravenously.
GMBF by CGRP was 45% of the pretreatment values (Figure 2B), despite a decrease in MABP by 12 mm Hg (Figure 2A). In the glibenclamide-treated group, GMBF increased by 16% during administration of CGRP (Figure 2B). The effect of CGRP on GMBF and MABP was almost absent in the CGRP-(8–37)–treated (500 pmol/min) group. In three respective vehicle control groups, MABP and GMBF were stable throughout the experiments (data not shown). Experiment 3 Effect of glibenclamide on gastric mucosal injury. Macroscopic examination. Intragastric perfusion with
0.15N HCl 1 15% ethanol induced hemorrhagic lesions in the glandular area of the stomach. The lesions in the glibenclamide group were larger than those in the vehicle control (DMSO) group but were smaller than those in the CGRP-(8–37) group (Figure 3).21 The lesions in the CGRP-(8–37) group and the glibenclamide 1 CGRP-(8– 37) group were comparable but were larger than those in the corresponding control groups (Figure 3). The lesions in the CGRP group were smaller than those in the corresponding vehicle control group (Figure 3). Microscopic examination. The proportion of section length occupied by deep damage was in the following ascending order: CGRP, control, glibenclamide, and CGRP-(8–37) groups (Figure 4). The percentages of
74
DOI ET AL.
GASTROENTEROLOGY Vol. 114, No. 1
Table 2. Physiological Variables in Experiment 2 CGRP Baseline PaCO2 (mm Hg) C 35.7 6 2.3 G1C 36.2 6 1.4 C(8–37) 1 C 34.6 6 2.0 PaO2 (mm Hg) C 80.5 6 5.7 G1C 81.8 6 8.2 C(8–37) 1 C 80.6 6 6.9 pH C 7.37 6 0.011 G1C 7.38 6 0.028 C(8–37) 1 C 7.36 6 0.029 Hematocrit (% ) C 44.5 6 1.4 G1C 42.8 6 2.8 C(8–37) 1 C 43.5 6 2.7 Glucose (mg/dL) C 100.4 6 11.6 G1C 104.6 6 22.6 C(8–37) 1 C 98.5 6 9.6
Pretreatment
After 5 min
36.2 6 1.7 36.5 6 1.7 33.1 6 2.5
34.4 6 3.7 35.1 6 4.2 32.2 6 2.2
83.5 6 1.4 83.5 6 1.4 81.2 6 4.7
84.2 6 5.4 a 84.2 6 5.4 a 84.1 6 7.1 a,b
7.39 6 0.096 7.39 6 0.031 7.38 6 0.24
7.43 6 0.051 a,b 7.43 6 0.104 a 7.42 6 0.49 a
45.7 6 1.7 44.0 6 4.8 44.5 6 3.7
44.2 6 3.4 43.0 6 3.7 42.9 6 4.9
98.0 6 13.9 90.1 6 7.9 c 97.9 6 11.5
108.1 6 13.0 88.9 6 23.8 a,c 100.8 6 16.9
NOTE. Values are means 6 SD. C, CGRP (20 pmol/min, intra-arterially); G, glibenclamide (20 mg/kg, intravenously); C(8–37), CGRP-(8–37) (500 pmol/min, intra-arterially). aP , 0.05 vs. baseline. bP , 0.05 vs. pretreatment. cP , 0.001 vs. CGRP and CGRP (8–37) 1 C.
lesion with shallow damage in CGRP, control, glibenclamide, and CGRP-(8–37) groups were 55.4% 6 11.1%, 40.5% 6 12.5%, 38.9% 6 19.2%, and 30.9% 6 24.3%, whereas those without detectable damage were 34.6% 6 6.6%, 30.1% 6 10.2%, 22.0% 6 15.6%, and 20.9% 6 25.3%, respectively.
Discussion There are three major findings in the present study. First, the activation of KATP channels increases GMBF. Second, a part of the increase in GMBF by CGRP is mediated by the activation of KATP channels. Finally, endogenous CGRP may protect the gastric mucosa against acid challenge, at least in part, through the activation of KATP channels. In some previous studies in which GMBF was measured by a hydrogen clearance method, GMBF was approximately 40 mL · min21 · 100 g tissue21 in rats.8,22 The value is lower than that reported in the present study. The inconsistency most likely results from the methodological difference. We used a contact-type electrode, whereas an insertion-type electrode was used in the previous studies. The value reported in the present study is similar to GMBF estimated by a contact-type electrode.17
Figure 2. Changes in (A) MABP and (B) GMBF in CGRP group (X; n 5 8), glibenclamide 1 CGRP group (N; n 5 8), and CGRP-(8–37) 1 CGRP group (S; n 5 6) in experiment 2. *P , 0.05 vs. glibenclamide 1 CGRP; †P , 0.05 vs. baseline and pretreatment in the same group; **P , 0.05 vs. CGRP-(8–37) 1 CGRP. Bars represent SD. ia, intra-arterially; iv, intravenously.
Intragastric perfusion with 0.15N HCl plus 15% ethanol produces gastric mucosal lesion histologically similar to human peptic ulcer. Furthermore, an increase in acid backdiffusion, which is known as one of the underlying mechanisms for peptic ulcer, is also reported in this model.14 Some investigations have been performed using this model to study mechanisms for gastric mucosal protection.10,14 These facts should validate the use of this model in the present experiments. Y-26763, CGRP, and CGRP-(8–37) were administered by a continuous infusion into the celiac artery. This procedure would be the best way to clarify the effect of these agents on the gastric mucosal circulation because any change in systemic hemodynamics such as blood pressure can be minimized with an intra-arterial injection. By contrast, glibenclamide was administered into the femoral vein as a bolus injection. This is based on previous reports that showed potent and long-lasting blockade of KATP channels with this procedure.23,24 Activation of KATP channels is an important mechanism for vasodilation.2,5 There are few studies, however, about the role of KATP channels in gastric circulation.25 In the present study, intra-arterial infusion of Y-26763 increased GMBF without changing systemic blood pres-
January 1998
Figure 3. Macroscopic gastric mucosal injury. Lesion Index (%) 5 (Lesion Area/Total Glandular Area) 3 100. *P , 0.05. Bars represent means 6 SD.
sure, and the responses were inhibited by glibenclamide, a selective inhibitor of KATP channels.26,27 These results suggest that KATP channels are present and functioning in gastric mucosal circulation. KATP channels play a role in the determination of resting tone in some3,4,28 but not all arteries or vascular beds.29 The present study showed that baseline GMBF was not affected by glibenclamide alone. This observation suggests that KATP channels do not contribute to the determination of baseline GMBF in Wistar rats. However, we should take the effect of anesthesia and surgical procedures into account. The involvement of KATP channels in vasodilation induced by CGRP is heterogeneous among different
Figure 4. Microscopic gastric mucosal injury. The vertical line represents the proportion of deep damage in examined sections. Deep damage refers to mucosal lesions affecting deeper than 25% of the mucosa. *P , 0.05. Bars represent means 6 SD. hCGRP-(8–37), human CGRR-(8–37).
ATP–SENSITIVE POTASSIUM CHANNEL AND CGRP
75
vascular beds.1,2,28 The increase in GMBF induced by CGRP, at the dose used in the present study, was inhibited by 64% in the presence of glibenclamide in rats. The result is consistent with the view that CGRP increases GMBF, at least in part, by activating KATP channels in the gastric vascular bed of rats. Because the increase in GMBF to CGRP was abolished by CGRP-(8– 37), the opening of KATP channels should be mediated by the activation of CGRP-1 receptors. Two major mechanisms by which CGRP activates KATP channels have been postulated: activation of adenylate cyclase in vascular smooth muscle2 and release of NO from the endothelium.30 It is possible that endothelial NO activates KATP channels on arterial smooth muscle, as shown in mesenteric arteries of rabbits.6 CGRP-induced increase in GMBF is believed to be an important defense mechanism of the gastric mucosa against ulcerogenic factors.10 Thus, KATP channels may also be involved in mucosal defense as an antiulcerogenic factor. In fact, glibenclamide and CGRP-(8–37) worsened gastric mucosal injuries in the present study. Furthermore, glibenclamide had no additional effects on CGRP-(8–37). This observation is consistent with the view that KATP channels are activated by CGRP and protect gastric mucosa during acid challenge. Thus, KATP channel openers may have a protective effect on gastric mucosa by increasing GMBF. The present study warrants further investigations for the clinical usefulness of KATP channel openers as an antiulcer drug.
References 1. Kitazono T, Heisted D, Faraci FM. Role of ATP-sensitive K1 channels in CGRP-induced dilation of basilar artery in vivo. Am J Physiol 1993;265:H581–H585. 2. Quayle MJ, Bonev AD, Brayden JE, Nelson MT. Calcitonin generelated peptide activated ATP-sensitive K1 currents in rabbit arterial smooth muscle via protein kinase A. J Physiol Lond 1994;475:9–13. 3. Nagao T, Sadoshima S, Kamouchi M, Fujishima M. Cromakalim dilates rat cerebral arteries in vivo. Stroke 1991;22:221–224. 4. Imamura Y, Tomoike H, Narishige T, Takahashi T, Kasuya H, Takeshita T. Glibenclamide decreases basal coronary blood flow in anesthetized dogs. Am J Physiol 1992;263:H399–H404. 5. Nelson MT, Huang Y, Brayden JE, Hescheler J, Standen NB. Arterial dilations in response to calcitonin gene-related peptide involve activation of K1 channels. Nature 1990;344:770–773. 6. Murphy EM, Brayden JE. Nitric oxide hyperpolarized rabbit mesenteric arteries via ATP-sensitive potassium channels. J Physiol Lond 1995;486:47–58. 7. Dart C, Standen NB. Adenosine-activated potassium current in smooth muscle cells isolated from the pig coronary artery. J Physiol Lond 1993;471:767–786. 8. Li DS, Raybould HE, Quintero E, Guth PH. Role of calcitonin gene-related peptide in gastric hyperemic response to intragastric capsaicin. Am J Physiol 1991;261:G657–G661. 9. Holzer P, Guth PH. Neuropeptide control of rat gastric mucosal blood flow: increase by calcitonin gene-related peptide and
76
10.
11.
12. 13. 14.
15.
16. 17.
18.
19.
20.
21.
DOI ET AL.
vasoactive intestinal peptide, but not substance P and neurokinin A. Circ Res 1991;68:100–105. Li DS, Raybould HE, Quintero E, Guth PH. Calcitonin gene-related peptide mediates the gastric hyperemic response to acid backdiffusion. Gastroenterology 1992;102:1124–1128. Zschauer A, Uusitalo H, Brayden JE. Role of endothelium and hyperpolarization in CGRP-induced vasodilation of rabbit ophthalmic artery. Am J Physiol 1992;263:H359–H365. Davenport HW. The gastric mucosal barrier. Digestion 1972;5: 162–165. Ritchie WP. Acute gastric mucosal damage induced by bile salts, acid and ischemia. Gastroenterology 1975;68:699–707. Holzer P, Livingston EH, Guth PH. Sensory neurons signal for an increase in rat gastric mucosal blood flow in the face of pending acid injury. Gastroenterology 1991;101:416–423. Chen R, Li DS, Guth PH. Capsaicin-sensitive afferent nerve induced gastric arteriolar dilatation is mediated in rat by calcitonin gene-related peptide (CGRP) (abstr). Gastroenterology 1991; 100:A42. Aukland K, Bower BF, Berliner RW. Measurement of local blood flow with hydrogen gas. Circ Res 1964;14:164–187. Murakami M, Morinaga M, Mitake T, Uchino H. Contact electrode method in hydrogen gas clearance technique: a new method for determination of regional gastric mucosal blood flow in animals and humans. Gastroenterology 1982;82:457–467. Kawakubo K, Ibayashi S, Nagao T, Doi K, Aoyagi K, Iida M, Sadoshima S, Fujishima M. Brain ischemia decreases gastric mucosal blood flow in hypertensive rats: role of arterial vagal adrenoceptors. Dig Dis Sci 1996;41:2383–2391. Nakajima T, Shinohara T, Yaoka O, Fukunari A, Shinagawa K, Aoki K, Katoh A, Yamanaka T, Setoguchi M, Tahara T. Y-27152, a long-acting K1 channel opener with less tachycardia: antihypertensive effects in hypertensive rats and dogs in conscious state. J Pharmacol Exp Ther 1992;261:730–736. Holzer P, Pabst MA, Lippe IT, Peskar BM, Peskar BA, Livingston EH, Guth PH. Afferent nerve-mediated protection against deep mucosal damage in the rat stomach. Gastroenterology 1990;98: 838–848. Doi K, Nagao T, Kawakubo K, Yano Y, Yamamoto C, Kanamoto K, Aoyagi K, Ibayashi T, Fujishima M. Calcitonin gene-related peptide protects against the formation of gastric mucosal lesions through the activation of ATP-sensitive potassium channel (abstr). Gastroenterology 1996;110:A96.
GASTROENTEROLOGY Vol. 114, No. 1
22. Chen YZR, Guth PH. Interaction of endogenous nitric oxide and CGRP in sensory neuron-induced gastric vasodilation. Am J Physiol 1995;268:791–796. 23. Thornton JD, Thornton CS, Sterling DL, Downey JM. Blockade of ATP-sensitive potassium channels increases infarct size but does not prevent preconditioning in rabbit hearts. Circ Res 1993;72: 44–49. 24. Furukawa S, Satoh K, Taira N. Opening of ATP-sensitive K1 channels responsible for adenosine A2 receptor–mediated vasodepression does not involve a pertussis toxin–sensitive G protein. Eur J Pharmacol 1993;236:255–262. 25. Cook NS, Fozard JR, Hof RP. The potential of potassium channel openers in peripheral vascular disease. In: Escande D, Standen N, eds. K1 channels in cardiovascular medicine. Paris: SpringerVerlag, 1993:305–319. 26. Ashford MLJ, Sturgess NC, Trout NJ, Gardner NJ, Hales CN. Adenosine 58-triphosphate–sensitive ion channels in neonatal rat culture central neurons. Pflugers Arch 1988;412:297–304. 27. Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT. Hyperpolarizing vasodilators activate ATP-sensitive K1 channels in arterial smooth muscle. Science 1989;245:177–180. 28. Nagao T, Ibayashi S, Sadoshima S, Fujii K, Fujii K, Ohya Y, Fujishima M. Distribution and physiological roles of ATP-sensitive K1 channels in the vertebrobasilar system of the rabbit. Circ Res 1996;78:238–243. 29. McPherson GA, Stork AP. The resistance of some rat cerebral arteries to the vasorelaxant effect of cromakalim and other K1 channel openers. Br J Pharmacol 1992;105:51–58. 30. Whittle BJR, Lopez-Belmonte J, Moncada S. Nitric oxide mediates rat mucosal vasodilatation induced by intragastric capsaicin. Eur J Pharmacol 1992;218:339–341.
Received October 30, 1996. Accepted September 16, 1997. Address requests for reprints to: Kosei Doi, M.D., Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka 812-82, Japan. Fax: (81) 92-642-5273. Presented in part at Digestive Disease Week in San Francisco, California, on May 19–22, 1996. The authors thank Yoshitomi Pharmaceuticals Ind., Ltd., Fukuoka, Japan, for providing Y-26763; and Fujisawa Pharmaceutical Co. Ltd., Osaka, Japan, for providing omeprazole.