- Email: [email protected]
The role of nitric oxide in the inhibition of gastric epithelial proliferation in portal hypertensive rats
Journal of Hepatology 1999; 301 1099-1104 Printed in Denmark All rights reserved Munksgaard Copenhagen
Copyright 8 European Association for the Study of the Liver 1999
Journal of Hepatology ISSN
0168-8278
The role of nitric oxide in the inhibition of gastric epithelial proliferation in portal hypertensive rats Dolgor Baatar, Seigo Kitano, Takanori
Yoshida, Toshio Bandoh,
Koichi Ninomiya
and Sadaki Tsuboi
First Department of Surgery, Oita Medical University, Japan
Background/Aim: Portal hypertension is associated with inhibition of gastric epithelial proliferation and increased gastric nitric oxide synthase activity. Whether the nitric oxide inhibits gastric epithelial proliferation is unclear. Metkfs: Portal vein ligation was performed to induce portal hypertension in rats. The rats were treated for 7 days with either vehicle or NG-nitro-L-arginine methyl ester (L-NAME) at 5 mg/kg or 25 mgkg doses (gastric gavage, twice a day). Sham-operated rats treated with vehicle served as controls. Hemodynamic parameters were measured using radiolabeled microspheres in anesthetized animals. Gastric epithelial proliferation was assessed by evaluating the proliferative cell nuclear antigen labeling index. Results: The cardiac index and gastric fundic blood flow were higher, and the gastric fundic proliferative cell nuclear antigen labeling index was lower in the portal hypertensive rats than in the controls. In portal
G
mucosal lesions have long been recognized in association with portal hypertension. These lesions, called “portal hypertensive gastropathy,” are now a well-established clinical entity (1). However, the pathogenesis of this disorder is still not clearly understood. Experimental studies have demonstrated that the portal hypertensive (PHT) gastric mucosa has several morphological and functional abnormalities and a relatively high susceptibility to injury by noxious agents (2). Recent studies have suggested that an excessive biosynthesis of the endothelium-derived vasoASTRIC
Received 15 September; revised 15 December; accepted 18 December 1998
Correspondence: Dolgor Baatar, First Department of Surgery, Oita Medical University, Oita 87995593, Japan. Tel: 81 97 5865843. Fax: 81 97 5496039. E-mail: baatareoita-med.ac.jp
hypertensive rats, the 5 mgkg dose of L-NAME decreased the cardiac index and increased the gastric fundic proliferative cell nuclear antigen labeling index to levels similar to those found in the controls, but did not affect gastric fundic blood flow significantly. The 25 mg/kg dose of L-NAME further decreased both the cardiac index and the gastric fundic blood flow, but did not affect the gastric proliferative cell nuclear antigen labeling index significantly. Conclusions: In portal hypertensive rats, the correction of systemic hyperdynamic circulation by NO inhibition is associated with normalization of gastric epithelial proliferation. Excessive nitric oxide may inhibit gastric epithelial proliferation in portal hypertension.
Key words: Gastric mucosa; Nitric oxide; Portal hypertension; Proliferating cell nuclear antigen.
dilator nitric oxide (NO) may be involved in the hyperdynamic circulation observed in portal hypertension (3,4). Increased gastric mucosal NO synthase activity has been demonstrated in patients with portal hypertensive gastropathy, suggesting an important role for NO in the pathogenesis of this mucosal lesion (5). Moreover, excessive gastric NO production has been implicated in the increased susceptibility to injury of PHT gastric mucosa (6). NO produced from NO donors inhibits the proliferation of different cell types (7,8). Portal hypertension in rats is associated with decreased gastric epithelial proliferation (9). Therefore, we hypothesized that the excessive NO may play a role in the inhibition of gastric epithelial proliferation that occurs in portal hypertension. In the present study, we investigated the effects of oral long-term administration of different doses of the NO synthesis inhibitor NG-nitro-L-arginine methyl ester (L-NAME) on 1099
D. Buatar et al
hemodynamics an animal
model
and
gastric
of portal
epithelial
proliferation
in
hypertension.
Materials and Methods Animals Procedures and protocols were reviewed and approved by the Animal Studies Committee at Oita Medical University. Sixty-four ‘i-week old male Sprague-Dawley rats were used in the present study. Portal hypertension was produced by staged portal vein ligation as described elsewhere (2). Sham operations were performed in control animals.
Study design Seven days after the first stage of portal vein ligation, portal hypertensive (PHT) rats were randomly assigned to receive either vehicle (distilled water) or L-NAME (Sigma Chemical Co., St. Louis, MO, USA) at doses of 5 mgikg body wt (L-NAME 5) or 25 mg/kg body wt (L-NAME 25) by gastric gavage twice a day. The last dose of LNAME or vehicle was given 12 h before the experiments. Sham-operated (SO) rats treated with vehicle served as controls. Hemodynamic and histological studies were performed after 7 days of treatment (i.e. 14 days after the first stage of portal vein ligation or sham operation).
Protocol 1: Nemodynamics In three groups of PHT rats treated with either vehicle, L-NAME 5 or L-NAME 25 and in one group of SO rats treated with vehicle (n= 8 in each group), cardiac output and gastric blood flow were measured using radioactive microspheres according to the reference sample technique (4,lO). After a 24-h fast with free access to water, each rat was anesthetized with pentobarbital sodium (Nembutal, Abbott Laboratories, North Chicago, IL, USA) at a dose of 50 mg/kg body wt, administered IF? Polyethylene catheters (PE 50, Becton Dickinson and Company. Parsippany, NJ, USA) were inserted into the left ventricle via the right carotid artery and into the abdominal aorta via the left femoral artery. The femoral artery line was connected to a pressure transducer (model TP-400T, Nihon Kohden. Tokyo, Japan), and the mean arterial blood pressure (MAP) was recorded on a polygraph system (RM-6000, Nihon Kohden), which was zero-referenced to the mid-chest of the animal and calibrated to mercury standards. Following a 30-min stabilization period, the MAP was registered and approximately 50 000 5’Cr-labeled microspheres (New England Nuclear, Boston, MA, USA), suspended in 0.1 ml of 0.02% Tween-80 (New England Nuclear), were injected into the left ventricle over 5 s. followed immediately by a 0.4-m] saline flash. Ten seconds before the microsphere injection, blood withdrawal was begun via the femoral artery catheter at a rate of 0.75 ml/min with a withdrawal pump (210, KD Scientific, Boston, MA, USA) and was continued for 1 min. After the rats were killed by pentobarbital sodium overdose (600 mg/kg body wt, IP), the stomachs were removed, opened along the greater curvature, and divided into the fore, fundic and antral sections, and weighed. The radioactivity of each sample was determined using a gamma counter (ARC-361, Aloka, Japan). The radioactivity of kidneys was also determined for the assessment of adequate mixing of microspheres (rats with a difference of more than 10% between the left and right kidney blood flows were excluded from the study). Cardiac output was calculated as cardiac output (ml/min)=(radioactivity injected [cpm]/reference blood sample radioactivity [cpm])x0.75 (ml/min). Cardiac output was expressed as cardiac index (CI) in ml/min per 100 g body wt (ml min-’ 100 g- ‘), Regional blood flows were calculated using the following formula: regional blood flow (ml/min)=(tissue sample radioactivity [cpm]/reference blood sample radoactivity[cpm])X0.75 (mlimin). Regional blood flows were expressed in mlimin per g tissue wt (ml. min-’ .g-‘).
Protocol 2: Gastric epithelial prokyeration Four groups of rats assigned to the same group categories used in Protocol 1 were studied. Each rat was anesthetized in the manner described above. Following midline laparotomy, the ileocolic vein was cannulated with a PE-50 catheter and the portal pressure (PP) was
1100
registered. Blood samples (2 ml) were obtained from the aorta following portal pressure measurements. Blood samples were centrifuged at 3000 rpm for 15 min at 4”C, and the separated serum was stored at -80°C until the time of assay. Serum gastrin concentrations were determined by radioimmunoassay using a Gastrin RIA Kit II (Dainabot, Tokyo, Japan). After the rats were killed. the stomachs were excised, opened along the greater curvature and fixed in methanol solution for 48 h at 4°C. Fixed gastric specimens were cut perpendicular to the gastric axis through the middle part of the stomach, which included both the fundic and antral mucosa. After routine processing and embedding in paraffin, five 3.pm-thick sections were cut from each specimen. Deparaffinized sections were stained with H&E or immunohistochemically stained by the enhanced polymer one-step (EPOS) staining method. as previously described (11). In brief, the sections were incubated with monoclonal mouse anti-proliferating cell nuclear antigen (PCNA) antibody (Dako EPOS U 07032. Dako Japan Co., Kyoto. Japan) for 1 h at room temperature. The color was developed with diaminobenzidine (DAB. Dako), and the sections were counterstained lightly with hematoxylin. For a control study. EPOS Negative Control (Dako) was used instead of primary antibody. Rat testes were used as a positive control tissue. A cell was regarded as labeled if the nucleus was stained distinctly brown. PCNA-labeled cells and the total number of cells in a 500,um-wide region of the total thickness of gastric mucosa were counted at a magnification of 400 with a square grid eyepiece graticule (Nikon, Tokyo, Japan). PCNA LI was expressed as the percentage of labeled cells/total cells. Measurements were made in two randomly chosen areas from both the fundic and antral mucosa (four in total). Four sections per rat were evaluated by two investigators under blind conditions. and the mean was calculated.
Additional stud) The effects of chronic treatment with L-NAME at a dose of 1 mg/kg body wt (L-NAME 1) given by gastric gavage twice a day for 7 days on hemodynamics (n=8) and gastric fundic epithelial proliferation (n=8) of the PHT rats was investigated in the same manner as described in Protocols I and 2.
Statistics All data are expressed as mean?SEM. The data were analyzed using a computer-based software system (StatView J-4.02. Abacus Concepts, Berkeley. CA. USA). Differences were tested by one-way analysis of variance (ANOVA). When ANOVA revealed significant differences f$
Results Hemodynamics The PP was markedly higher in the vehicle-treated PHT rats than in the SO rats (Table 1). In PHT rats, L-NAME 1 and L-NAME 5 did not affect the PP significantly, whereas L-NAME 25 even increased it compared with PP following administration of vehicle. The MAP was lower and the CI was higher in the vehicletreated PHT rats than in the SO animals. PHT rats treated with L-NAME 1 and L-NAME 5 had MAP and CI levels similar to those registered in SO rats. PHT rats treated with L-NAME 25 had significantly higher MAP and lower CI than those in either the SO rats or the PHT rats treated with L-NAME 1 and LNAME 5. Gastric fundic blood flow was significantly higher in
NO and proliferation TABLE
of PHT gastric mucosa
1
Hemodynamic response to chronic treatment NAME 5) and 25 mg/kg body wt (L-NAME
with vehicle (VHC) or L-NAME at doses of 1 mg/kg body wt (L-NAME 25) in sham-operated (SO) and portal hypertensive (PHT) rats Portal
pressure
Group
lmmHg1
1mnWl
Cardiac index [ml . min-’ . 100 g-r]
SO+VHC PHT+VHC PHT+L-NAME PHT+L-NAME PHT+L-NAME
5.7520.16 11.38?0.49* 11.86?0.74* 11.29?0.78* 15.33t1.12*+t
115.50-r-2.03 96.87+3.48* 117.29+2.86+ 117.71-C3.29+ 143.50?3.65*+*
21.63kO.95 28.38+2.33* 21.92?1.58+ 19.24-c 1.29+ 13.57*0.47*+x
Data are mean+-SEM, 5 (Scheffe’s F-test).
n=7-8
1 5 25
in each group.
Mean arterial
*p
+p
the vehicle-treated PHT rats than in the SO rats (Fig. 1). In PHT rats, L-NAME tended to decrease the fundic blood flow dose-dependently Fundic blood flow in the PHT rats treated with L-NAME 1 was significantly higher than that in the SO rats and similar to that in the vehicle-treated PHT rats. Fundic blood flow in the PHT rats treated with L-NAME 5 was similar to that in the SO rats, but still was not significantly different from that in the vehicle-treated PHT rats. In PHT rats, L-NAME 25 significantly decreased fundic blood flow relative to vehicle and relative to L-NAME 1. The antral blood flow did not differ significantly between the vehicle-treated PHT rats and the SO rats. In PHT rats, treatment with L-NAME 5 and LNAME 25 did not affect the antral blood flow significantly. Gastric epithelial proliferation In fundic mucosa, positive reaction with anti-PCNA antibody was observed in the nucleus of cells localized
Fundus
WC
J
I
L-PLME
1
I.-M
I
5 PHT
L-NAME
25,
Antrum
I
I
WC
L.-J SO
I
5 _..PH I
pressure
2s ,
Fig. I. Gastric fundic and antral blood flows of sham-operated (SO) andportal hypertensive (PHT) rats chronically treated with either vehicle (VHC) or L-NAME at doses of 1 mg/kg (L-NAME I), 5 mg/kg (L-NAME 5) and 25 mg/ kg (L-NAME 25). Data are mean?SEM, n=7-8 in each group. *p
$<0.05
vs PHT+L-NAME
l), 5 mg/kg
body wt (L-
1 and PHT+L-NAME
predominantly in the mucous neck region, which corresponds to a morphologically identified proliferative zone (Fig. 2). Many cells in the proliferative zone were stained positively in a fundic mucosa section obtained from an SO rat treated with vehicle (Fig. 2a). In contrast, there were only a few positively stained cells in a section from a PHT rat treated with vehicle (Fig. 2b). In the fundic mucosa of a PHT rat treated with L-NAME 5, the number of positively stained cells was similar to that found in an SO rat (Fig. 2~). The mean fundic PCNA LI in the vehicle-treated PHT rats was 11.27?0.46%, which was significantly less than the index of 16.53? 1.05% in the SO rats (Fig. 3). In PHT rats, both L-NAME 1 and L-NAME 5 increased fundic PCNA LI to levels similar to that registered in SO rats, whereas L-NAME 25 did not affect it significantly. The fundic PCNA LI in the PHT rats treated with L-NAME 25 was significantly lower than those in the SO rats and the PHT rats treated with L-NAME 1 and L-NAME 5. In the antral mucosa, the PCNA LI did not differ significantly between the SO rats and the PHT rats treated with vehicle. Compared to vehicle, L-NAME 5 and L-NAME 25 did not affect PCNA LI in the antral mucosa of PHT animals. The mean fasting serum gastrin levels did not differ significantly between the four groups: 154.86+ 15.54 pg/ml in the SO rats, 136.30~21.68 pg/ml in the PHT rats treated with vehicle, 184.012 37.15 pg/ml in the PHT rats treated with L-NAME 5 and 168.98221.74 pg/ml in the PHT rats treated with L-NAME 25 @= 0.5922, one-way ANOVA).
Discussion In the present study, a significant increase in CI and decrease in MAP were registered in PHT rats, demonstrating the development of portal hypertension with systemic hyperdynamic circulation. Moreover, our PHT rats had decreased gastric fundic PCNA LI, suggesting that gastric epithelial proliferation was inhib1101
D. Baatar et al.
Fig. 2. Microphotographs of gastric fundic mucosal sections immunostained with anti-prohferuting body of vehicle-treated (A) sham-operated and (B) portal hypertensive rats, and (C) L=NAME hypertensive rut (origin& magntfiication X.50, bar= 100 pm).
ited in this model of portal hypertension. Chronic administration of the NO inhibitor at the dose which corrected MAP and CI, increased gastric fundic PCNA LI in PHT rats to the control levels. These results suggest that excessive NO may play a role not only in the development of systemic hyperdynamic circulation, but also in the inhibition of gastric epithelial proliferation in PHT rats. A previous study showed that L-NAME, at a dose normalizing arterial vasodilatation and hypotension in
Fundus
1
+
!---I
SO
I
5
PHT
d
r-- -
L--IL_
SO
5
2s ,
PHT
Fig. 3. Proltferuting cell nuclear antigen (PCNA) labeling index (LI) in fundic and antral mucosa of shum-operated (SO) and portal hypertensive (PHT) rats chronically treated either with vehicle ( VHC) or L-NAME at doses of I mg/kg (L-NAME I), 5 mg/‘kg (L-NAME 5) and 25 mg/ kg (L-NAME 25). Data me meanLSEM, n=8 in each group. *p
cell nuclear antigen anti(5 my/kg)-treated portal
cirrhotic rats, decreased NO production (aortic cyclic GMP) to control levels (3). Our PHT rats treated with L-NAME 1 and L-NAME 5 had systemic hemodynamic parameters similar to those observed in the control rats, suggesting that these doses of L-NAME may decrease NO production in PHT animals to levels close to normal. Regarding splanchnic hemodynamics, we found that gastric blood flow is high in PHT rats mainly due to the increase in fundic blood flow. Total gastric blood flow and gastric mucosal blood flow have been shown to be increased in the different models of portal hypertension (4,6,12,13). Various factors, such as NO and prostaglandins, are involved in the gastric hyperemic response associated with portal hypertension (4,6,13). Inhibition of NO synthase with a single 5 mg/kg dose of L-NAME normalizes increased NO generation and mucosal blood flow in the stomachs of PHT rats (6). In our study, chronic treatment with LNAME 1 did not significantly affect, and treatment with L-NAME 5 only slightly decreased, gastric fundic blood flow in the PHT rats. In PHT rats, the vasoconstrictor effect of chronic NO biosynthesis inhibition on splanchnic circulation, including gastric blood flow, is attenuated, while its systemic effect is maintained (4). It has been suggested that the splanchnic production of prostaglandins may be increased after NO inhibition, counteracting the L-NAME-induced vasoconstriction. Accordingly, it is possible that enhanced gastric prostaglandin synthesis attenuated the effects of chronic NO inhibition on gastric blood flow in our PHT rats. Chronic NO inhibition in PHT rats also increases porto-collateral resistance (4). Increased porto-collateral
NO and proliferation
resistance, together with attenuated vasoconstrictor effect on splanchnic circulation may explain the absence of portal hypotensive effect of chronic NO inhibition in PHT rats. In the present study, we used the PCNA LI to estimate the proliferative activity of the gastric epithelia. The expression of PCNA is necessary for DNA synthesis, and this expression is maximal during the S phase of the cell cycle (14). PCNA immunohistochemistry is a reliable marker of cell proliferation in the gastrointestinal tract of the rats (15). Therefore, we believe that in our study, the registered changes in PCNA LI reflect corresponding alterations in gastric epithelial proliferation. In PHT rats, L-NAME increased gastric fundic epithelial proliferation only at those doses which did not significantly decrease gastric fundic blood flow. The suppression of gastric mucosal blood flow by inhibition of endogenous NO is associated with a decrease in gastric mucosal DNA synthesis (16). The release of endogenous NO and the consequent vasodilatation have been implicated in the stimulation of gastric mucosal growth induced by sensory nerves. In PHT rats, treatment with L-NAME 25 markedly decreased gastric fundic blood flow, suggesting that this dose of LNAME may suppress gastric NO production to a level which may alter the above-mentioned mechanism controlling gastric mucosal growth. In our study, portal hypertension inhibited epithelial proliferation in fundic mucosa but not in antral mucosa. A previous study showed that epithelial proliferation was decreased in both the fundic and the antral areas of PHT rats (9). This discrepancy between the previous findings and ours may be explained by the fact that we performed our study a shorter time after the portal vein ligation (2 weeks), than that used in the previous study (2 months). Prolonged portal hypertension may inhibit epithelial proliferation in total gastric mucosa. Portal hypertensive gastropathy is predominantly located in the fundus, and it has been suggested that the corpus of the stomach may be more susceptible than the antrum to congestive mucosal circulation caused by portal hypertension (17). In our study, only the fundic blood flow was increased significantly in PHT rats. Furthermore, L-NAME 25 decreased fundic blood flow but did not affect antral blood flow significantly. These results suggest that NO production in portal hypertension may increase primarily in the fundic area, thus contributing to the increase in fundic blood flow and the inhibition of fundic epithelial proliferation in PHT rats. The study aimed to determine gastric mucosal NO synthase activity separately in the
of PHT gastric mucosa
gastric fundic and antral areas of PHT rats may help to prove this hypothesis. The mechanisms by which excessive production of gastric NO can inhibit gastric epithelial proliferation remain to be elucidated. Gastric epithelial proliferation is regulated by several factors, such as gastrin and the growth factors of the epidermal growth factor (EGF) family (18,19). In the present study, fasting gastrin levels did not differ significantly between PHT and SO rats. Furthermore, NO inhibition in PHT rats was not associated with significant alterations in serum gastrin levels, suggesting that gastrin-related mechanisms are not involved in the changes in gastric epithelial proliferation registered in our study. EGF family members exert their effects by activating the intrinsic tyrosine kinase of their common EGF receptor (20). NO released from NO donor drugs causes a reversible inhibition of the EGF receptor tyrosine kinase activity of tibroblasts, which overexpress the human EGF receptor (7). Moreover, NO donors inhibit EGF-induced DNA synthesis and subsequent proliferation of vascular smooth muscle cells (8). Therefore, it is possible that excessive gastric NO may inhibit EGF-induced proliferation of gastric epithelial cells. In summary, the present study has shown that the correction of systemic hyperdynamic circulation by long-term NO-synthesis inhibition is associated with a normalization of gastric epithelial proliferation in PHT rats. Thus, an overproduction of NO seems to be involved in the inhibition of gastric epithelial proliferation in portal hypertension,
References 1. McCormack TT, Sims J, Eyre-Brook I, Kennedy H, Goepel J, Johnson AG, et al. Gastric lesions in portal hypertension: inflammatory gastritis or congestive gastropathy? Gut 1985; 26: 1226 32. 2. Sarpheh IJ, Tarnawski A. Increased susceptibility of the portal hypertensive gastric mucosa to damage. J Clin Gastroenterol 1991; 13: Sl8-21. 3. Niederberger M, Martin PY, Gines P Morris K, Tsai P, Xu DL, et al. Normalization of nitric oxide production corrects arterial vasodilatation and hyperdynamic circulation in cirrhotic rats. Gastroenterology 1995; 109: 1624-30. 4. Garcia-Pagan JC, Fernandez M, Bernadich C, Pizcueta P Pique JM, Bosch J, et al. Effects of continued NO inhibition on portal hypertensive syndrome after portal vein stenosis. Am J Physiol 1994; 267: G984-90. 5. El-Newihi HM, Kanji VK, Mihas AA. Activity of gastric mucosal nitric oxide synthase in portal hypertensive gastropathy. Am J Gastroenterol 1996; 91: 535-8. 6. Ohta M, Tanoue K, Tarnawski A, Pai R, Itani RM, Sander FC, et al. Overexpressed nitric oxide synthase in portal-hypertensive stomach of rat: a key to increased susceptibility to damage? Gastroenterology 1997; 112: 192&30. 7. Estrada C, Gomez C, Martin-Nieto J, De Frutos T, Jimenez A, Villalobo A. Nitric oxide reversibly inhibits the epidermal growth factor receptor tyrosine kinase. Biochem J 1997; 326: 369-76. 8. Yu SM, Hung LM, Lin CC. cGMP-elevating agents suppress
1103
D. Baatar et al.
9. 10.
11.
12.
13.
14.
proliferation of vascular smooth muscle cells by inhibiting the activation of epidermal growth factor signaling pathway. Circulation 1997; 95: 1269977. Hsieh JS, Huang TJ. The effect of portal hypertension on gastric epithelial proliferation in rats. Eur Sur Res 1993; 25: 98103. Heymann MA, Payne BD, Hoffman JI. Rudolph AM. Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Res 1977; 20: 55-79. Tsutsumi Y, Serizawa A, Kawai K. Enhanced polymer one-step staining (EPOS) for proliferating cell nuclear antigen (PCNA) and Ki-67 antigen. Pathol Int 1995; 45: 108-15. Kitano S, Koyonagi N, Sugimdchi K, Kobayashi M, Inokuchi K. Mucosal blood flow and modified vascular responses to norepinephrine in the stomach of rats with liver cirrhosis. Em Sur Res 1982; 14: 221-30. Pique JM, Marroni N, Bosch J, Whittle BJ. Involvement of nitric oxide and prostaglandins in gastric mucosal hyperemia of portal hypertensive anesthetized rats. Hepatology 1993; 18: 628-34. Celis JE, Celis A. Cell cycle-dependant variations in the distribution of the nuclear protein cyclin proliferating cell nuclear anti-
1104
15.
16.
17.
18.
19.
20.
gen in cultured cell: subdivision of S phase. Proc Nat1 Acad Sci USA 1985; 82: 3262-6. Yamada K, Yoshitake, Sato M, Ahnen DJ. Proliferating cell nuclear antigen expression in normal. preneoplastic, and neoplastic colonic epithelium of the rat. Gastroenterology 1992; 103: 160-7. Dembinski A, Warzecha Z, Ceranowicz P Konturek SJ. The role of capsaicin-sensory neurons and nitric oxide in regulation of gastric mucosal growth. J Physiol Pharm 1995; 46: 351-62. Iwao T, Toyonaga A. lkegami M, Oho K, Sumino M. Harada H. et al. Reduced gastric mucosal blood flow in patients with portal hypertensive gastropathy. Hepatology 1993; 18: 3640. Ohning GV, Wong HC. Lloyd KC, Walsh JH. Gastrin mediates the gastric mucosal proliferative response to feeding. Am J Phisi01 1996; 271: G470-6. Johnson LR, Guthrie PD. Stimulation of rat oxyntic gland mucosal growth by epidermal growth factor. Am J Physiol 1980; 238: G45-9. Ullrich A, Schlessinger J. Signal transduction by receptors with tyrosine kinase activity. Cell 1990; 20: 203312.
Copyright 8 European Association for the Study of the Liver 1999
Journal of Hepatology ISSN
0168-8278
The role of nitric oxide in the inhibition of gastric epithelial proliferation in portal hypertensive rats Dolgor Baatar, Seigo Kitano, Takanori
Yoshida, Toshio Bandoh,
Koichi Ninomiya
and Sadaki Tsuboi
First Department of Surgery, Oita Medical University, Japan
Background/Aim: Portal hypertension is associated with inhibition of gastric epithelial proliferation and increased gastric nitric oxide synthase activity. Whether the nitric oxide inhibits gastric epithelial proliferation is unclear. Metkfs: Portal vein ligation was performed to induce portal hypertension in rats. The rats were treated for 7 days with either vehicle or NG-nitro-L-arginine methyl ester (L-NAME) at 5 mg/kg or 25 mgkg doses (gastric gavage, twice a day). Sham-operated rats treated with vehicle served as controls. Hemodynamic parameters were measured using radiolabeled microspheres in anesthetized animals. Gastric epithelial proliferation was assessed by evaluating the proliferative cell nuclear antigen labeling index. Results: The cardiac index and gastric fundic blood flow were higher, and the gastric fundic proliferative cell nuclear antigen labeling index was lower in the portal hypertensive rats than in the controls. In portal
G
mucosal lesions have long been recognized in association with portal hypertension. These lesions, called “portal hypertensive gastropathy,” are now a well-established clinical entity (1). However, the pathogenesis of this disorder is still not clearly understood. Experimental studies have demonstrated that the portal hypertensive (PHT) gastric mucosa has several morphological and functional abnormalities and a relatively high susceptibility to injury by noxious agents (2). Recent studies have suggested that an excessive biosynthesis of the endothelium-derived vasoASTRIC
Received 15 September; revised 15 December; accepted 18 December 1998
Correspondence: Dolgor Baatar, First Department of Surgery, Oita Medical University, Oita 87995593, Japan. Tel: 81 97 5865843. Fax: 81 97 5496039. E-mail: baatareoita-med.ac.jp
hypertensive rats, the 5 mgkg dose of L-NAME decreased the cardiac index and increased the gastric fundic proliferative cell nuclear antigen labeling index to levels similar to those found in the controls, but did not affect gastric fundic blood flow significantly. The 25 mg/kg dose of L-NAME further decreased both the cardiac index and the gastric fundic blood flow, but did not affect the gastric proliferative cell nuclear antigen labeling index significantly. Conclusions: In portal hypertensive rats, the correction of systemic hyperdynamic circulation by NO inhibition is associated with normalization of gastric epithelial proliferation. Excessive nitric oxide may inhibit gastric epithelial proliferation in portal hypertension.
Key words: Gastric mucosa; Nitric oxide; Portal hypertension; Proliferating cell nuclear antigen.
dilator nitric oxide (NO) may be involved in the hyperdynamic circulation observed in portal hypertension (3,4). Increased gastric mucosal NO synthase activity has been demonstrated in patients with portal hypertensive gastropathy, suggesting an important role for NO in the pathogenesis of this mucosal lesion (5). Moreover, excessive gastric NO production has been implicated in the increased susceptibility to injury of PHT gastric mucosa (6). NO produced from NO donors inhibits the proliferation of different cell types (7,8). Portal hypertension in rats is associated with decreased gastric epithelial proliferation (9). Therefore, we hypothesized that the excessive NO may play a role in the inhibition of gastric epithelial proliferation that occurs in portal hypertension. In the present study, we investigated the effects of oral long-term administration of different doses of the NO synthesis inhibitor NG-nitro-L-arginine methyl ester (L-NAME) on 1099
D. Buatar et al
hemodynamics an animal
model
and
gastric
of portal
epithelial
proliferation
in
hypertension.
Materials and Methods Animals Procedures and protocols were reviewed and approved by the Animal Studies Committee at Oita Medical University. Sixty-four ‘i-week old male Sprague-Dawley rats were used in the present study. Portal hypertension was produced by staged portal vein ligation as described elsewhere (2). Sham operations were performed in control animals.
Study design Seven days after the first stage of portal vein ligation, portal hypertensive (PHT) rats were randomly assigned to receive either vehicle (distilled water) or L-NAME (Sigma Chemical Co., St. Louis, MO, USA) at doses of 5 mgikg body wt (L-NAME 5) or 25 mg/kg body wt (L-NAME 25) by gastric gavage twice a day. The last dose of LNAME or vehicle was given 12 h before the experiments. Sham-operated (SO) rats treated with vehicle served as controls. Hemodynamic and histological studies were performed after 7 days of treatment (i.e. 14 days after the first stage of portal vein ligation or sham operation).
Protocol 1: Nemodynamics In three groups of PHT rats treated with either vehicle, L-NAME 5 or L-NAME 25 and in one group of SO rats treated with vehicle (n= 8 in each group), cardiac output and gastric blood flow were measured using radioactive microspheres according to the reference sample technique (4,lO). After a 24-h fast with free access to water, each rat was anesthetized with pentobarbital sodium (Nembutal, Abbott Laboratories, North Chicago, IL, USA) at a dose of 50 mg/kg body wt, administered IF? Polyethylene catheters (PE 50, Becton Dickinson and Company. Parsippany, NJ, USA) were inserted into the left ventricle via the right carotid artery and into the abdominal aorta via the left femoral artery. The femoral artery line was connected to a pressure transducer (model TP-400T, Nihon Kohden. Tokyo, Japan), and the mean arterial blood pressure (MAP) was recorded on a polygraph system (RM-6000, Nihon Kohden), which was zero-referenced to the mid-chest of the animal and calibrated to mercury standards. Following a 30-min stabilization period, the MAP was registered and approximately 50 000 5’Cr-labeled microspheres (New England Nuclear, Boston, MA, USA), suspended in 0.1 ml of 0.02% Tween-80 (New England Nuclear), were injected into the left ventricle over 5 s. followed immediately by a 0.4-m] saline flash. Ten seconds before the microsphere injection, blood withdrawal was begun via the femoral artery catheter at a rate of 0.75 ml/min with a withdrawal pump (210, KD Scientific, Boston, MA, USA) and was continued for 1 min. After the rats were killed by pentobarbital sodium overdose (600 mg/kg body wt, IP), the stomachs were removed, opened along the greater curvature, and divided into the fore, fundic and antral sections, and weighed. The radioactivity of each sample was determined using a gamma counter (ARC-361, Aloka, Japan). The radioactivity of kidneys was also determined for the assessment of adequate mixing of microspheres (rats with a difference of more than 10% between the left and right kidney blood flows were excluded from the study). Cardiac output was calculated as cardiac output (ml/min)=(radioactivity injected [cpm]/reference blood sample radioactivity [cpm])x0.75 (ml/min). Cardiac output was expressed as cardiac index (CI) in ml/min per 100 g body wt (ml min-’ 100 g- ‘), Regional blood flows were calculated using the following formula: regional blood flow (ml/min)=(tissue sample radioactivity [cpm]/reference blood sample radoactivity[cpm])X0.75 (mlimin). Regional blood flows were expressed in mlimin per g tissue wt (ml. min-’ .g-‘).
Protocol 2: Gastric epithelial prokyeration Four groups of rats assigned to the same group categories used in Protocol 1 were studied. Each rat was anesthetized in the manner described above. Following midline laparotomy, the ileocolic vein was cannulated with a PE-50 catheter and the portal pressure (PP) was
1100
registered. Blood samples (2 ml) were obtained from the aorta following portal pressure measurements. Blood samples were centrifuged at 3000 rpm for 15 min at 4”C, and the separated serum was stored at -80°C until the time of assay. Serum gastrin concentrations were determined by radioimmunoassay using a Gastrin RIA Kit II (Dainabot, Tokyo, Japan). After the rats were killed. the stomachs were excised, opened along the greater curvature and fixed in methanol solution for 48 h at 4°C. Fixed gastric specimens were cut perpendicular to the gastric axis through the middle part of the stomach, which included both the fundic and antral mucosa. After routine processing and embedding in paraffin, five 3.pm-thick sections were cut from each specimen. Deparaffinized sections were stained with H&E or immunohistochemically stained by the enhanced polymer one-step (EPOS) staining method. as previously described (11). In brief, the sections were incubated with monoclonal mouse anti-proliferating cell nuclear antigen (PCNA) antibody (Dako EPOS U 07032. Dako Japan Co., Kyoto. Japan) for 1 h at room temperature. The color was developed with diaminobenzidine (DAB. Dako), and the sections were counterstained lightly with hematoxylin. For a control study. EPOS Negative Control (Dako) was used instead of primary antibody. Rat testes were used as a positive control tissue. A cell was regarded as labeled if the nucleus was stained distinctly brown. PCNA-labeled cells and the total number of cells in a 500,um-wide region of the total thickness of gastric mucosa were counted at a magnification of 400 with a square grid eyepiece graticule (Nikon, Tokyo, Japan). PCNA LI was expressed as the percentage of labeled cells/total cells. Measurements were made in two randomly chosen areas from both the fundic and antral mucosa (four in total). Four sections per rat were evaluated by two investigators under blind conditions. and the mean was calculated.
Additional stud) The effects of chronic treatment with L-NAME at a dose of 1 mg/kg body wt (L-NAME 1) given by gastric gavage twice a day for 7 days on hemodynamics (n=8) and gastric fundic epithelial proliferation (n=8) of the PHT rats was investigated in the same manner as described in Protocols I and 2.
Statistics All data are expressed as mean?SEM. The data were analyzed using a computer-based software system (StatView J-4.02. Abacus Concepts, Berkeley. CA. USA). Differences were tested by one-way analysis of variance (ANOVA). When ANOVA revealed significant differences f$
Results Hemodynamics The PP was markedly higher in the vehicle-treated PHT rats than in the SO rats (Table 1). In PHT rats, L-NAME 1 and L-NAME 5 did not affect the PP significantly, whereas L-NAME 25 even increased it compared with PP following administration of vehicle. The MAP was lower and the CI was higher in the vehicletreated PHT rats than in the SO animals. PHT rats treated with L-NAME 1 and L-NAME 5 had MAP and CI levels similar to those registered in SO rats. PHT rats treated with L-NAME 25 had significantly higher MAP and lower CI than those in either the SO rats or the PHT rats treated with L-NAME 1 and LNAME 5. Gastric fundic blood flow was significantly higher in
NO and proliferation TABLE
of PHT gastric mucosa
1
Hemodynamic response to chronic treatment NAME 5) and 25 mg/kg body wt (L-NAME
with vehicle (VHC) or L-NAME at doses of 1 mg/kg body wt (L-NAME 25) in sham-operated (SO) and portal hypertensive (PHT) rats Portal
pressure
Group
lmmHg1
1mnWl
Cardiac index [ml . min-’ . 100 g-r]
SO+VHC PHT+VHC PHT+L-NAME PHT+L-NAME PHT+L-NAME
5.7520.16 11.38?0.49* 11.86?0.74* 11.29?0.78* 15.33t1.12*+t
115.50-r-2.03 96.87+3.48* 117.29+2.86+ 117.71-C3.29+ 143.50?3.65*+*
21.63kO.95 28.38+2.33* 21.92?1.58+ 19.24-c 1.29+ 13.57*0.47*+x
Data are mean+-SEM, 5 (Scheffe’s F-test).
n=7-8
1 5 25
in each group.
Mean arterial
*p
+p
the vehicle-treated PHT rats than in the SO rats (Fig. 1). In PHT rats, L-NAME tended to decrease the fundic blood flow dose-dependently Fundic blood flow in the PHT rats treated with L-NAME 1 was significantly higher than that in the SO rats and similar to that in the vehicle-treated PHT rats. Fundic blood flow in the PHT rats treated with L-NAME 5 was similar to that in the SO rats, but still was not significantly different from that in the vehicle-treated PHT rats. In PHT rats, L-NAME 25 significantly decreased fundic blood flow relative to vehicle and relative to L-NAME 1. The antral blood flow did not differ significantly between the vehicle-treated PHT rats and the SO rats. In PHT rats, treatment with L-NAME 5 and LNAME 25 did not affect the antral blood flow significantly. Gastric epithelial proliferation In fundic mucosa, positive reaction with anti-PCNA antibody was observed in the nucleus of cells localized
Fundus
WC
J
I
L-PLME
1
I.-M
I
5 PHT
L-NAME
25,
Antrum
I
I
WC
L.-J SO
I
5 _..PH I
pressure
2s ,
Fig. I. Gastric fundic and antral blood flows of sham-operated (SO) andportal hypertensive (PHT) rats chronically treated with either vehicle (VHC) or L-NAME at doses of 1 mg/kg (L-NAME I), 5 mg/kg (L-NAME 5) and 25 mg/ kg (L-NAME 25). Data are mean?SEM, n=7-8 in each group. *p
$<0.05
vs PHT+L-NAME
l), 5 mg/kg
body wt (L-
1 and PHT+L-NAME
predominantly in the mucous neck region, which corresponds to a morphologically identified proliferative zone (Fig. 2). Many cells in the proliferative zone were stained positively in a fundic mucosa section obtained from an SO rat treated with vehicle (Fig. 2a). In contrast, there were only a few positively stained cells in a section from a PHT rat treated with vehicle (Fig. 2b). In the fundic mucosa of a PHT rat treated with L-NAME 5, the number of positively stained cells was similar to that found in an SO rat (Fig. 2~). The mean fundic PCNA LI in the vehicle-treated PHT rats was 11.27?0.46%, which was significantly less than the index of 16.53? 1.05% in the SO rats (Fig. 3). In PHT rats, both L-NAME 1 and L-NAME 5 increased fundic PCNA LI to levels similar to that registered in SO rats, whereas L-NAME 25 did not affect it significantly. The fundic PCNA LI in the PHT rats treated with L-NAME 25 was significantly lower than those in the SO rats and the PHT rats treated with L-NAME 1 and L-NAME 5. In the antral mucosa, the PCNA LI did not differ significantly between the SO rats and the PHT rats treated with vehicle. Compared to vehicle, L-NAME 5 and L-NAME 25 did not affect PCNA LI in the antral mucosa of PHT animals. The mean fasting serum gastrin levels did not differ significantly between the four groups: 154.86+ 15.54 pg/ml in the SO rats, 136.30~21.68 pg/ml in the PHT rats treated with vehicle, 184.012 37.15 pg/ml in the PHT rats treated with L-NAME 5 and 168.98221.74 pg/ml in the PHT rats treated with L-NAME 25 @= 0.5922, one-way ANOVA).
Discussion In the present study, a significant increase in CI and decrease in MAP were registered in PHT rats, demonstrating the development of portal hypertension with systemic hyperdynamic circulation. Moreover, our PHT rats had decreased gastric fundic PCNA LI, suggesting that gastric epithelial proliferation was inhib1101
D. Baatar et al.
Fig. 2. Microphotographs of gastric fundic mucosal sections immunostained with anti-prohferuting body of vehicle-treated (A) sham-operated and (B) portal hypertensive rats, and (C) L=NAME hypertensive rut (origin& magntfiication X.50, bar= 100 pm).
ited in this model of portal hypertension. Chronic administration of the NO inhibitor at the dose which corrected MAP and CI, increased gastric fundic PCNA LI in PHT rats to the control levels. These results suggest that excessive NO may play a role not only in the development of systemic hyperdynamic circulation, but also in the inhibition of gastric epithelial proliferation in PHT rats. A previous study showed that L-NAME, at a dose normalizing arterial vasodilatation and hypotension in
Fundus
1
+
!---I
SO
I
5
PHT
d
r-- -
L--IL_
SO
5
2s ,
PHT
Fig. 3. Proltferuting cell nuclear antigen (PCNA) labeling index (LI) in fundic and antral mucosa of shum-operated (SO) and portal hypertensive (PHT) rats chronically treated either with vehicle ( VHC) or L-NAME at doses of I mg/kg (L-NAME I), 5 mg/‘kg (L-NAME 5) and 25 mg/ kg (L-NAME 25). Data me meanLSEM, n=8 in each group. *p
cell nuclear antigen anti(5 my/kg)-treated portal
cirrhotic rats, decreased NO production (aortic cyclic GMP) to control levels (3). Our PHT rats treated with L-NAME 1 and L-NAME 5 had systemic hemodynamic parameters similar to those observed in the control rats, suggesting that these doses of L-NAME may decrease NO production in PHT animals to levels close to normal. Regarding splanchnic hemodynamics, we found that gastric blood flow is high in PHT rats mainly due to the increase in fundic blood flow. Total gastric blood flow and gastric mucosal blood flow have been shown to be increased in the different models of portal hypertension (4,6,12,13). Various factors, such as NO and prostaglandins, are involved in the gastric hyperemic response associated with portal hypertension (4,6,13). Inhibition of NO synthase with a single 5 mg/kg dose of L-NAME normalizes increased NO generation and mucosal blood flow in the stomachs of PHT rats (6). In our study, chronic treatment with LNAME 1 did not significantly affect, and treatment with L-NAME 5 only slightly decreased, gastric fundic blood flow in the PHT rats. In PHT rats, the vasoconstrictor effect of chronic NO biosynthesis inhibition on splanchnic circulation, including gastric blood flow, is attenuated, while its systemic effect is maintained (4). It has been suggested that the splanchnic production of prostaglandins may be increased after NO inhibition, counteracting the L-NAME-induced vasoconstriction. Accordingly, it is possible that enhanced gastric prostaglandin synthesis attenuated the effects of chronic NO inhibition on gastric blood flow in our PHT rats. Chronic NO inhibition in PHT rats also increases porto-collateral resistance (4). Increased porto-collateral
NO and proliferation
resistance, together with attenuated vasoconstrictor effect on splanchnic circulation may explain the absence of portal hypotensive effect of chronic NO inhibition in PHT rats. In the present study, we used the PCNA LI to estimate the proliferative activity of the gastric epithelia. The expression of PCNA is necessary for DNA synthesis, and this expression is maximal during the S phase of the cell cycle (14). PCNA immunohistochemistry is a reliable marker of cell proliferation in the gastrointestinal tract of the rats (15). Therefore, we believe that in our study, the registered changes in PCNA LI reflect corresponding alterations in gastric epithelial proliferation. In PHT rats, L-NAME increased gastric fundic epithelial proliferation only at those doses which did not significantly decrease gastric fundic blood flow. The suppression of gastric mucosal blood flow by inhibition of endogenous NO is associated with a decrease in gastric mucosal DNA synthesis (16). The release of endogenous NO and the consequent vasodilatation have been implicated in the stimulation of gastric mucosal growth induced by sensory nerves. In PHT rats, treatment with L-NAME 25 markedly decreased gastric fundic blood flow, suggesting that this dose of LNAME may suppress gastric NO production to a level which may alter the above-mentioned mechanism controlling gastric mucosal growth. In our study, portal hypertension inhibited epithelial proliferation in fundic mucosa but not in antral mucosa. A previous study showed that epithelial proliferation was decreased in both the fundic and the antral areas of PHT rats (9). This discrepancy between the previous findings and ours may be explained by the fact that we performed our study a shorter time after the portal vein ligation (2 weeks), than that used in the previous study (2 months). Prolonged portal hypertension may inhibit epithelial proliferation in total gastric mucosa. Portal hypertensive gastropathy is predominantly located in the fundus, and it has been suggested that the corpus of the stomach may be more susceptible than the antrum to congestive mucosal circulation caused by portal hypertension (17). In our study, only the fundic blood flow was increased significantly in PHT rats. Furthermore, L-NAME 25 decreased fundic blood flow but did not affect antral blood flow significantly. These results suggest that NO production in portal hypertension may increase primarily in the fundic area, thus contributing to the increase in fundic blood flow and the inhibition of fundic epithelial proliferation in PHT rats. The study aimed to determine gastric mucosal NO synthase activity separately in the
of PHT gastric mucosa
gastric fundic and antral areas of PHT rats may help to prove this hypothesis. The mechanisms by which excessive production of gastric NO can inhibit gastric epithelial proliferation remain to be elucidated. Gastric epithelial proliferation is regulated by several factors, such as gastrin and the growth factors of the epidermal growth factor (EGF) family (18,19). In the present study, fasting gastrin levels did not differ significantly between PHT and SO rats. Furthermore, NO inhibition in PHT rats was not associated with significant alterations in serum gastrin levels, suggesting that gastrin-related mechanisms are not involved in the changes in gastric epithelial proliferation registered in our study. EGF family members exert their effects by activating the intrinsic tyrosine kinase of their common EGF receptor (20). NO released from NO donor drugs causes a reversible inhibition of the EGF receptor tyrosine kinase activity of tibroblasts, which overexpress the human EGF receptor (7). Moreover, NO donors inhibit EGF-induced DNA synthesis and subsequent proliferation of vascular smooth muscle cells (8). Therefore, it is possible that excessive gastric NO may inhibit EGF-induced proliferation of gastric epithelial cells. In summary, the present study has shown that the correction of systemic hyperdynamic circulation by long-term NO-synthesis inhibition is associated with a normalization of gastric epithelial proliferation in PHT rats. Thus, an overproduction of NO seems to be involved in the inhibition of gastric epithelial proliferation in portal hypertension,
References 1. McCormack TT, Sims J, Eyre-Brook I, Kennedy H, Goepel J, Johnson AG, et al. Gastric lesions in portal hypertension: inflammatory gastritis or congestive gastropathy? Gut 1985; 26: 1226 32. 2. Sarpheh IJ, Tarnawski A. Increased susceptibility of the portal hypertensive gastric mucosa to damage. J Clin Gastroenterol 1991; 13: Sl8-21. 3. Niederberger M, Martin PY, Gines P Morris K, Tsai P, Xu DL, et al. Normalization of nitric oxide production corrects arterial vasodilatation and hyperdynamic circulation in cirrhotic rats. Gastroenterology 1995; 109: 1624-30. 4. Garcia-Pagan JC, Fernandez M, Bernadich C, Pizcueta P Pique JM, Bosch J, et al. Effects of continued NO inhibition on portal hypertensive syndrome after portal vein stenosis. Am J Physiol 1994; 267: G984-90. 5. El-Newihi HM, Kanji VK, Mihas AA. Activity of gastric mucosal nitric oxide synthase in portal hypertensive gastropathy. Am J Gastroenterol 1996; 91: 535-8. 6. Ohta M, Tanoue K, Tarnawski A, Pai R, Itani RM, Sander FC, et al. Overexpressed nitric oxide synthase in portal-hypertensive stomach of rat: a key to increased susceptibility to damage? Gastroenterology 1997; 112: 192&30. 7. Estrada C, Gomez C, Martin-Nieto J, De Frutos T, Jimenez A, Villalobo A. Nitric oxide reversibly inhibits the epidermal growth factor receptor tyrosine kinase. Biochem J 1997; 326: 369-76. 8. Yu SM, Hung LM, Lin CC. cGMP-elevating agents suppress
1103
D. Baatar et al.
9. 10.
11.
12.
13.
14.
proliferation of vascular smooth muscle cells by inhibiting the activation of epidermal growth factor signaling pathway. Circulation 1997; 95: 1269977. Hsieh JS, Huang TJ. The effect of portal hypertension on gastric epithelial proliferation in rats. Eur Sur Res 1993; 25: 98103. Heymann MA, Payne BD, Hoffman JI. Rudolph AM. Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Res 1977; 20: 55-79. Tsutsumi Y, Serizawa A, Kawai K. Enhanced polymer one-step staining (EPOS) for proliferating cell nuclear antigen (PCNA) and Ki-67 antigen. Pathol Int 1995; 45: 108-15. Kitano S, Koyonagi N, Sugimdchi K, Kobayashi M, Inokuchi K. Mucosal blood flow and modified vascular responses to norepinephrine in the stomach of rats with liver cirrhosis. Em Sur Res 1982; 14: 221-30. Pique JM, Marroni N, Bosch J, Whittle BJ. Involvement of nitric oxide and prostaglandins in gastric mucosal hyperemia of portal hypertensive anesthetized rats. Hepatology 1993; 18: 628-34. Celis JE, Celis A. Cell cycle-dependant variations in the distribution of the nuclear protein cyclin proliferating cell nuclear anti-
1104
15.
16.
17.
18.
19.
20.
gen in cultured cell: subdivision of S phase. Proc Nat1 Acad Sci USA 1985; 82: 3262-6. Yamada K, Yoshitake, Sato M, Ahnen DJ. Proliferating cell nuclear antigen expression in normal. preneoplastic, and neoplastic colonic epithelium of the rat. Gastroenterology 1992; 103: 160-7. Dembinski A, Warzecha Z, Ceranowicz P Konturek SJ. The role of capsaicin-sensory neurons and nitric oxide in regulation of gastric mucosal growth. J Physiol Pharm 1995; 46: 351-62. Iwao T, Toyonaga A. lkegami M, Oho K, Sumino M. Harada H. et al. Reduced gastric mucosal blood flow in patients with portal hypertensive gastropathy. Hepatology 1993; 18: 3640. Ohning GV, Wong HC. Lloyd KC, Walsh JH. Gastrin mediates the gastric mucosal proliferative response to feeding. Am J Phisi01 1996; 271: G470-6. Johnson LR, Guthrie PD. Stimulation of rat oxyntic gland mucosal growth by epidermal growth factor. Am J Physiol 1980; 238: G45-9. Ullrich A, Schlessinger J. Signal transduction by receptors with tyrosine kinase activity. Cell 1990; 20: 203312.