Effects of potent bombesin antagonist on exocrine pancreatic secretion in rats.

Recent synthesis of specific, potent bombesin receptor antagonists allows examination of the role of bombesin-like peptides in physiological processes in vivo. We characterized effects of [D-Phe6]bombesin(6-13)-methyl-ester (BME) on pancreatic enzyme secretion stimulated by the C-terminal decapeptide of gastrin releasing peptide (GRP-10), food intake, and diversion of bile-pancreatic juice in rats. In isolated pancreatic acini, BME had no agonistic effects on amylase secretion but competitively inhibited responses to GRP-10, yielding a pA2 value of 8.89 +/- 0.19. In conscious rats with gastric, jugular vein, bile-pancreatic, and duodenal cannulas, basal enzyme secretion (bile-pancreatic juice recirculated) was not affected by the antagonist. Maximal amylase response to GRP-10 (0.5 nmol/kg/h) was inhibited dose dependently by BME, reaching 97% inhibition at a dose of 400 nmol/kg/h. The dose response curve of amylase secretion stimulated by GRP-10 was shifted to the right by 40 nmol/kg/h BME, but maximal amylase response was unaltered, suggesting competitive inhibition in vivo. Liquid food intake and bile-pancreatic juice diversion caused substantial increases in amylase secretion; neither response was altered during administration of 400 pmol/kg/h BME. These results demonstrate that BME is a potent, competitive antagonist of pancreatic responses to bombesin-like peptides in vitro and in vivo. Lack of effect of BME on basal pancreatic secretion or responses to liquid food intake or diversion of bile-pancreatic juice in rats suggests that endogenous bombesin-like peptides do not act either directly or indirectly to mediate these responses.     

Elevation of plasma CCK concentration after intestinal administration of a pancreatic enzyme secretion-stimulating peptide purified from rat bile-pancreatic juice: analysis with N-terminal region specific radioimmunoassay

The rat plasma cholecystokinin (CCK) concentration was measured after intestinal administration of a peptide purified from rat bile-pancreatic juice, which has a stimulatory effect on pancreatic enzyme secretion. The plasma CCK concentration was measured by means of a radioimmunoassay using CCK-8 N-terminal specific antibody, OAL-656. In experimental rats with protease-free intestines, intraduodenal infusion of 10 micrograms of the purified peptide, which stimulates pancreatic enzyme secretion 2.0-2.5 fold, induced a significant increase in the plasma CCK level. Furthermore, after removal of CCK from the plasma by immunoabsorption with an OAL-656-bound Sepharose 4B column, the stimulatory effect of the plasma on pancreatic enzyme secretion was abolished when it was injected intravenously into recipient rats. It was concluded that this peptide stimulates the release of CCK in the intestine and that this is responsible at least in part for the pancreatic enzyme secretion-stimulating activity of the peptide.     

The effect of oleate on pancreatic and bile secretion in the conscious rat

The effects of sodium oleate infused into either the duodenum or the terminal ileum on bile and pancreatic secretion were examined in the conscious rat. Rats were prepared with cannulae draining pure bile and pancreatic juice separately, and with an ileal and two duodenal cannulae. A 40 mM taurocholate solution containing 7 mg/ml bovine trypsin was infused into the duodenum throughout the experiment to replace diverted bile-pancreatic juice to maintain the normal regulation of pancreatic secretion. The intraduodenal infusion of sodium oleate significantly increased pancreatic juice flow, protein, and bicarbonate outputs, whereas it did not affect bile secretion. Intravenous infusion of proglumide (300 mg/kg/hr) did not inhibit pancreatic secretion stimulated by intraduodenal infusion of sodium oleate. An intravenous infusion of atropine (100 micrograms/kg/hr) attenuated protein and fluid secretions but not that of bicarbonate in response to intraduodenal oleate. In contrast, the intraileal infusion of oleate had no effect on pancreatic secretion, whereas it decreased bile flow, bicarbonate, and bile salt outputs. In conclusion, sodium oleate introduced in the duodenum stimulates pancreatic secretion but oleate in the terminal ileum inhibits bile secretion.     

Inhibitory effect of pancreastatin on pancreatic exocrine secretion in the conscious rat

The effect of newly discovered pancreastatin on pancreatic secretion stimulated by a diversion of bile-pancreatic juice (BPJ) from the intestine was examined in the conscious rat. Exogenous pancreastatin infusion (20, 100 and 200 pmol/kg.h) inhibited pancreatic protein and fluid outputs during BPJ diversion in a dose-dependent manner. Pancreastatin did not affect plasma cholecystokinin (CCK) concentrations. Pancreastatin (100 pmol/kg.h) inhibited CCK-stimulated pancreatic secretion, but did not inhibit secretin-stimulated pancreatic secretion. Pancreastatin alone, however, did not affect basal pancreatic secretion. In contrast, pancreastatin (10(-10)-10(-7)M) did not suppress CCK-stimulated amylase release from isolated rat pancreatic acini. These results indicate that pancreastatin has an inhibitory action on exocrine function of the pancreas. This action may not be mediated by direct mechanisms and nor via an inhibition of CCK release. It is suggested that pancreastatin may play a role in the regulation of the intestinal phase of exocrine pancreatic secretion.     

Effect of diversion of bile-pancreatic juice to the ileum on pancreatic secretion and adaptation in the rat

Cannulas were implanted to collect bile and pancreatic juice, and the collected secretions were pumped back into the intestine at the level of the duodenum or the proximal ileum. The effect of 6 days of such treatment on pancreatic secretion and on pancreatic growth was determined. The effect on pancreatic secretion was studied by measuring the pancreatic secretory response to a stimulus, provided by acute diversion of bile-pancreatic juice from the proximal intestine. Trophic effects were studied in a separate group of rats by measuring pancreatic weight, protein content, and chymotrypsin activity after an overnight fast. Stimulated pancreatic secretion was 2.1 times greater for protein output and 3.4 times greater for fluid output in rats with chronic diversion of bile-pancreatic juice to the ileum. Pancreatic weight, protein content, and chymotrypsin activity were increased 2.6, 2.9, and 4.8 times, respectively, by chronic diversion of bile-pancreatic juice to the ileum. These results indicate that pancreatic hypertrophy and hyperplasia reported in rats with bile-pancreatic duct transposition to the ileum are the result of loss of feed-back inhibition from bile-pancreatic juice in the proximal intestine.     

The effect of luminal ghrelin on pancreatic enzyme secretion in the rat

Ghrelin, a 28-amino-acid peptide produced predominantly by oxyntic mucosa has been reported to affect the pancreatic exocrine function but the mechanism of its secretory action is not clear. The effects of intraduodenal (i.d.) infusion of ghrelin on pancreatic amylase outputs under basal conditions and following the stimulation of pancreatic secretion with diversion of pancreato-biliary juice (DPBJ) as well as the role of vagal nerve, sensory fibers and CCK in this process were determined. Ghrelin given into the duodenum of healthy rats at doses of 1.0 or 10.0 microg/kg increased pancreatic amylase outputs under basal conditions or following the stimulation of pancreatic secretion with DPBJ. Bilateral vagotomy as well as capsaicin deactivation of sensory fibers completely abolished all stimulatory effects of luminal ghrelin on pancreatic exocrine function. Pretreatment with lorglumide, a CCK(1) receptor blocker, reversed the stimulation of amylase release produced by intraduodenal application of ghrelin. Intraduodenal ghrelin at doses of 1.0 or 10.0 microg/kg increased plasma concentrations of CCK and ghrelin. In conclusion, ghrelin given into the duodenum stimulates pancreatic enzyme secretion. Activation of vagal reflexes and CCK release as well as central mechanisms could be implicated in the stimulatory effect of luminal ghrelin on the pancreatic exocrine functions.     

Nervous system regulation of exocrine pancreatic secretion in the chicken

Acute assays were carried out using broiler chickens in which a reentry catheter had previously been placed chronically in the main pancreatic duct. Samples of pancreatic juice were collected after manoeuvres of blockage and stimulation with different neurotransmitters and blocking agents, both cholinergic and adrenergic, as well as electrical stimulation of the left vagosympathetic trunk. Stimulation of the vagus nerve induced a marked increase in the pancreatic flow but with no changes in the enzyme content. Acetylcholine was seen to cause a slight but significant increase in pancreatic flow and a non-significant increase in amylase activity. The drop in the flow of pancreatic juice in response to adrenaline was not very sensitive to adrenergic blockers. The effect of adrenaline on pancreatic secretion cannot be attributed to vascular changes.     

Guanidinated casein hydrolysate stimulation of cholecystokinin release via pancreatic enzyme- and cholinergic-independent mechanisms in rats.

We had demonstrated that a peptic hydrolysate of guanidinated casein that is made from casein by the conversion of lysine to homoarginine stimulated pancreatic exocrine secretion in rats with chronic bile-pancreatic juice (BPJ) diversion from the proximal small intestine. This modified protein also stimulated cholecystokinin (CCK) release from dispersed rat intestinal cells. In this study, we found that guanidinated casein hydrolysate stimulates CCK release in chronic BPJ-diverted rats with cholinergic control blocked by atropine. Intraduodenal guanidinated casein hydrolysate increased portal plasma CCK concentration and pancreatic secretion in atropine-treated BPJ-diverted rats. In contrast, the portal plasma CCK concentration was not increased by intact casein hydrolysate. We conclude that guanidinated casein hydrolysate directly stimulates CCK release from the intestine via some cholinergic-independent mechanism, and an increase of the pancreatic exocrine secretion is regulated by CCK released by guanidinated casein hydrolysate. A guanidyl residue is likely to be involved in this control.     

Role of cholecystokinin in the intestinal fat- and acid-induced inhibition of gastric secretion.

This study was designed to determine the role of cholecystokinin (CCK) in the inhibition of gastric HCl secretion by duodenal peptone, fat and acid in dogs with chronic gastric and pancreatic fistulas. Intraduodenal instillation of 5% peptone stimulated both gastric HCl secretion and pancreatic protein secretion and caused significant increments in plasma gastrin and CCK levels. L-364,718, a selective antagonist of CCK-A receptors, caused further increase in gastric HCl and plasma gastrin responses to duodenal peptone but reduced the pancreatic protein outputs in these tests by about 75%. L-365,260, an antagonist of type B receptors, reduced gastric acid by about 25% but failed to influence pancreatic response to duodenal peptone. Addition of 10% oleate or acidification of peptone to pH 3.0 profoundly inhibited acid secretion while significantly increasing the pancreatic protein secretion and plasma CCK levels. Administration of L-364,718 reversed the fall in gastric HCl secretion and significantly attenuated pancreatic protein secretion in tests with both peptone plus oleate and peptone plus acid. Exogenous CCK infused i.v. in a dose (25 pmol/kg per h) that raised plasma CCK to the level similar to that achieved by peptone meal plus fat resulted in similar inhibition of gastric acid response to that attained with fat and this effect was completely abolished by the pretreatment with L-364,718. We conclude that CCK released by intestinal peptone meal, containing fat or acid, exerts a tonic inhibitory influence on gastric acid secretion and gastrin release through the CCK-A receptors.     

Stimulatory effect of an endogenous peptide in rat pancreatic juice on pancreatic enzyme secretion in the presence of atropine: evidence for different mode of action of stimulation from exogenous trypsin inhibitors

A new factor which activated the secretion of pancreatic enzymes was discovered and purified from rat bile-pancreatic juice. A fraction below M.W.10,000 of rat bile-pancreatic juice enhanced trypsinogen secretion by injection into anesthetized rat duodenum. The factor was purified from this fraction using its biological activity as an index by Sephadex G-50, SP Sephadex C-50 and HPLC. This factor was a peptide of which molecular weight was about 6,000 and had trypsin inhibitory activity. From these and some other findings, it was suggested that the peptide was identical with the "Kazal type" inhibitor. In the anesthetized and atropine-treated rat, of which intestinal trypsin was removed by thoroughly washing with saline containing 5 microM soybean trypsin inhibitor (SBTI), pancreatic secretion became basal state, and was not stimulated by injection of SBTI into its duodenum any longer. Under this condition, however, injection of this purified peptide brought about markedly stimulation of pancreatic enzyme secretion. These results suggest that this peptide has a certain function which enhances pancreatic enzyme secretion by the different manner from exogenous trypsin inhibitors such as SBTI.     

Atrial natriuretic factor stimulates exocrine pancreatic secretion in the rat through NPR-C receptors

Increasing evidence supports the role of atrial natriuretic factor (ANF) in the modulation of gastrointestinal physiology. The effect of ANF on exocrine pancreatic secretion and the possible receptors and pathways involved were studied in vivo. Anesthetized rats were prepared with pancreatic duct cannulation, pyloric ligation, and bile diversion into the duodenum. ANF dose-dependently increased pancreatic secretion of fluid and proteins and enhanced secretin and CCK-evoked response. ANF decreased chloride secretion and increased the pH of the pancreatic juice. Neither cholinergic nor adrenergic blockade affected ANF-stimulated pancreatic secretion. Furthermore, ANF response was not mediated by the release of nitric oxide. ANF-evoked protein secretion was not inhibited by truncal vagotomy, atropine, or Nomega-nitro-l-arginine methyl ester administration. The selective natriuretic peptide receptor-C (NPR-C) receptor agonist cANP-(4-23) mimicked ANF response in a dose-dependent fashion. When the intracellular signaling coupled to NPR-C receptors was investigated in isolated pancreatic acini, results showed that ANF did not modify basal or forskolin-evoked cAMP formation, but it dose-dependently enhanced phosphoinositide hydrolysis, which was blocked by the selective PLC inhibitor U-73122. ANF stimulated exocrine pancreatic secretion in the rat, and its effect was not mediated by nitric oxide or parasympathetic or sympathetic activity. Furthermore, CCK and secretin appear not to be involved in ANF response. Present findings support that ANF exerts a stimulatory effect on pancreatic exocrine secretion mediated by NPR-C receptors coupled to the phosphoinositide pathway.     

Responses of the exocrine pancreatic secretion to spontaneous feeding in rats with bile-pancreatic juice diversion

The regulatory response of the exocrine pancreas was examined in rats under unanesthetized and unrestrained conditions. The previous study demonstrated that the pancreatic protease secretion increased 2-fold after spontaneous feeding of a low protein diet in chronically bile-pancreatic cannulated rats (normal rats) whose bile-pancreatic juice (BPJ) was returned to the duodenum. In the present study, we observed the response of the exocrine pancreatic secretion to spontaneous feeding of a low protein diet in rats with chronic diversion of BPJ from the proximal small intestine for 6 days (bypass rat) whose diverted BPJ was returned to the upper ileum. During BPJ diversion, the dry weight and the protein content of the pancreas were increased 2-fold, compared with normal rats. Also, the levels of trypsinogen and chymotrypsinogen in the pancreas were increased several times, but amylase was decreased. The basal secretion of enzymes after a 24-hr fast was enhanced in bypass rats in proportion to the pancreatic enzyme contents. After spontaneous feeding of 8% casein fat-free diet, the increases in the pancreatic secretion of bypass rats were much smaller than those of normal rats. In contrast, the increase of BPJ flow of bypass rats after feeding was greater than that of normal rats. These findings represent that the chronic diversion of BPJ exerts hypergrowth of pancreas and hypersecretion of proteases in the fasting state, and less sensitivity of pancreatic enzyme secretion to dietary feeding.     

Caerulein-induced acute pancreatitis in the rat. Pancreatic secretory response to cholecystokinin.

The response of pancreatic exocrine secretion to cholecystokinin (CCK), has been studied in experimental acute pancreatitis induced in rats by supramaximal doses of caerulein. Several doses of caerulein were used (4, 20 and 40 micrograms/Kg) and each one was administered by four subcutaneous injections over 3 h at hourly intervals. Pancreatic juice was collected 9 h after the first injection. The caerulein-treated animals showed a statistically significant increase in serum amylase levels. Secretory activity of ductular cells remained unchanged in all the caerulein-treated animals, but total protein and amylase secretion decreased significantly at all the caerulein doses used, both in resting conditions and under stimulation with CCK (1.25 micrograms/Kg/h). Despite this the acinar cells of rats treated with the lowest dose of caerulein retained a certain degree of secretory function since amylase activity in pancreatic juice was greater than in other groups of rats treated with higher doses of caerulein. Moreover, the percentage of increase observed in total protein and amylase in response to CCK respect to basal secretion is similar to that of the untreated animals. At higher doses (20 and 40 micrograms/Kg) the secretory capacity in response to CCK was inhibited. Therefore CCK administration in slight acute pancreatitis could be used as a therapy since it favours the secretion of pancreatic enzymes at percentual levels similar to those of the controls.     

Interactions between bile and pancreatic juice diversions on cholecystokinin release and pancreas in conscious rats

Pancreatic exocrine secretion in the conscious rat is regulated by proteases secreted by the pancreas, and cholecystokinin (CCK) is known to be involved in its mechanism. It has also been reported that the absence of either pancreatic juice or bile in the duodenum could stimulate pancreatic secretion. Therefore, differences in CCK release responding to the exclusion of bile, pancreatic juice (PJ), or both bile and pancreatic juice (BPJ) from the intestine were examined by using a bioassay for cholecystokinin. Plasma CCK levels were increased by all three treatments compared with the basal value, the order of their effects being BPJ greater than PJ greater than bile diversion, and CCK concentrations produced by BPJ diversion were much greater than can be explained as simply summed effect of exclusions of bile and PJ. Pancreatic exocrine secretions were significantly increased by PJ and BPJ diversions, but the effect of bile diversion on the pancreas was not statistically significant. An additional infusion of CR-1409 (0.1 mg/rat), one of CCK receptor antagonists, inhibited exocrine function stimulated by BPJ diversion. We conclude (i) BPJ diversion is the strongest endogenous stimulant on CCK release; (ii) the potentiation between bile and PJ diversions is induced on CCK release; (iii) pancreatic protein secretion during BPJ diversion is mainly modulated by CCK.     

Factors involved in the intestinal feedback regulation of pancreatic enzyme secretion in the rat.

Further studies on the feedback regulation of pancreatic enzyme secretion by trypsin were conducted in conscious rats, surgically prepared so that pancreatic juice could be collected or returned. Suppression of enzyme secretion by trypsin as well as its stimulation by SBTI occurred only in the upper part of the small intestine, where the hormone CCK is known to be released. Over a limited range, trypsin suppression of pancreatic secretion was proportional to the dose of trypsin. Higher concentrations had no further effect, suggesting "saturation" of the intestine. Trypsin which had its active center blocked by DFP did not suppress enzyme output. These results supported the concept that only trypsin (or chymotrypsin) with an exposed active center suppressed pancreatic enzyme secretion in the rat by somehow suppressing the release of CCK from the intestinal cell. Presumably CCK is released from the intestine following "removal" of trypsin from the intestine either by diverting the juice or by feeding SBTI which binds the enzyme. All of the evidence supported the view that the effect of trypsin or SBTI on pancreatic secretion was mediated at the intestinal level and not in the blood as has been suggested.     

Competition of a growth stimulating-/cholecystokinin (CCK) releasing-peptide (monitor peptide) with epidermal growth factor for binding to 3T3 fibroblasts

The growth stimulating-/cholecystokinin (CCK) releasing-peptide (monitor peptide) is a peptide purified from rat bile-pancreatic juice on the basis of its stimulatory activity toward pancreatic enzyme secretion. Its multiple functions and peptide sequence suggested that it is distinct from epidermal growth factor (EGF). However, we found that the peptide competes with [125I]-EGF in the binding to Swiss 3T3 fibroblast cells to almost the same extent as unlabeled EGF does. [125I]-EGF binding was inhibited by 50% by the peptide at 82.8 ng/ml and by unlabeled EGF at 71.4 ng/ml. This suggests that the growth stimulating effect of the peptide on 3T3 fibroblasts is mediated via the EGF receptor, and also suggests that the partial homologous sequence between monitor peptide and EGF is required for the receptor binding, or that the EGF receptor has a broad ligand specificity.     

Appearance of New Protein Species on the Cell Surface during Sexual Differentiation in the Heterobasidiomycetous Yeast,Tremella mesenterica

The effects of neural blockers on the pancreatic enzyme secretion in response to an intraluminal infusion of soybean trypsin inhibitor and HCl were investigated. The stimulation of pancreatic enzyme secretion upon the intraluminal infusion of soybean trypsin inhibitor was not blocked by atropine, but was completely blocked by guanethidine. The intraluminal infusion of 0.08 n HCl, which is known as a potent secretagogue of secretin, caused a rapid augmentation of trypsin output, which was not blocked by atropine or guanethidine. Preinjection of CR-1392 (1.5 mg/kg, i.p.), which is a strong cholecystokinin receptor antagonist, completely blocked the pancreatic response to soybean trypsin inhibitor, but not that to 0.08 n HCl. This inferred that guanethidine specifically suppressed the CCK-release from the small intestine.

These findings suggest that the pancreatic enzyme secretion in response to soybean trypsin inhibitor is mainly mediated by CCK, and that adrenergic modulation would be involved in the CCK-mediated pancreatic enzyme secretion in response to soybean trypsin inhibitor.     

Pancreatic secretory response to feeding in the calf: CCK-A receptors, but not CCK-B/gastrin receptors are involved

In bovine species, as in human, the pancreas predominantly expresses cholecystokinin-B (CCK-B)/gastrin receptors. However, the role of this receptor in the regulation of meal-stimulated pancreatic enzyme release has not been determined. In milk-fed calves, we previously described prandial patterns of exocrine pancreatic secretion and a long prefeeding phase was observed. The present study was aimed at determining both the role of external stimuli in the outset of the prefeeding phase and the implication of pancreatic CCK-A and CCK-B/gastrin receptors in the mediation of pancreatic response to feeding. The first objective was studied by suppressing external stimuli associated with food intake (unexpected meal) and the second by infusing highly specific and potent antagonists of CCK-A (SR 27897) and CCK-B/gastrin (PD 135158) receptors during the prandial period. When calves were given an unexpected meal, the long prefeeding increase in pancreatic secretion was absent. SR 27897 (but not PD 135158) inhibited the preprandial phase and greatly reduced postprandial pancreatic juice and enzyme outflows. The expectancy of a meal seemed to elicit an increased pancreatic response right before a meal and CCK-A receptors may mediate this information via neural pathways. The implication of CCK and CCK-A receptors in mediating the postfeeding pancreatic response was also demonstrated. The participation of CCK-B/gastrin receptors in this regulation was not demonstrated.     

Bovine pancreatic secretion in the first week of life: potential involvement of intestinal CCK receptors.

The aim of this study was to evaluate pancreatic juice secretion of calves in the first postnatal days, and determine a potential involvement of cholecystokinin (CCK) and intestinal CCK receptor in its regulation. Nine neonatal Friesian calves (five controls and four treated intraduodenally with FK480, a CCK-A receptor antagonist) were surgically fitted with a pancreatic duct catheter and a duodenal cannula before the first colostrum feeding. Collections of pancreatic juice and duodenal luminal pressure recordings were started early after recovery from anaesthesia and continued for 6 days. From day 2 or 3 of life, periodic fluctuations in pancreatic secretions were observed in concert with duodenal myoelectric motor complex (MMC) and variations in plasma pancreatic polypeptide (PP) concentrations. Intraduodenal administration of FK480 reduced pancreatic juice secretion while intravenous infusion of CCK had no effect. Immunocytochemistry indicated an association of mucosal CCK-A and -B receptors with neural components of the small intestine. In conclusion, periodic activity of the exocrine pancreas exists in neonatal calves soon after birth and local neural intestinal CCK-A receptors could be partly responsible for the modulation of neonatal calf pancreatic secretion.     

<Emphasis Type="SmallCaps">l</Emphasis>-Homoarginine suppresses exocrine pancreas in rats

Summary.  Previously, we found that guanidinated casein, a l-homoarginine-containing protein, was a more potent stimulator of pancreatic enzyme secretion than intact casein in rats. In this study, we examined secretory response and adaptation of the exocrine pancreas to the administration of free l-homoarginine in normal and bile-pancreatic juice (BPJ)-diverted rats. An intraperitoneal injection of l-homoarginine (10 mg/rats) produced immediate and transient reduction in pancreatic secretion in BPJ-diverted rats, but not in normal rats. The BPJ-diverted rats were fed with either a 25% casein, 45% casein, or 45% casein diet supplemented with l-homoarginine (19 g/kg diet) for 4 days. Feeding of a diet containing l-homoarginine inhibited the pancreatic adaptation induced by the high-protein diet. These results indicate that l-homoarginine has an inhibitory effect on the secretion and production of exocrine pancreatic enzyme in BPJ-diverted rats, and l-homoarginine may have an antagonistic effect on CCK receptors. Received July 1, 2002 Accepted August 28, 2002 Published online December 20, 2002 Authors' address: Dr. Hiroshi Hara, Division of Applied Bioscience, Graduate School of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan E-mail: hara@chem.agr.hokudai.ac.jp