Hepatic vagal amino acid sensors modulate amino acid induced insulin and glucagon secretion in the rat

Journal of the Autonomic Nen~ous System, 42 (1993) 225-232

225

© 1993 Elsevier Science Publishers B.V. All rights reserved 0165-1838/93/$06.00 JANS 01366

Hepatic vagal amino acid sensors modulate amino acid induced insulin and glucagon secretion in the rat Katsuaki T a n a k a , Shuji I n o u e , S a t o r u Saito, Hajime Nagase and Yutaro T a k a m u r a The Third Department of Internal Medicine, Yokohama City Unit,ersity School of Medicine, Fukuura, Yokohama, Japan (Received 3 August 1992) (Revision received 22 September 1992) (Accepted 24 September 1992)

Key words: Hepatic vagotomy; Celiac vagotomy; Amino acids; Insulin; Glucagon Abstract To clarify the physiological role of vagal amino acid sensors in the liver, the effect of hepatic vagotomy and/or celiac vagotomy (sectioning of the hepatic branch and/or the celiac branches of the vagus nerve) on the secretion of insulin and glucagon after intraperitoneal injection of neutral (L-alanine, L-leucine, and L-phenylalanine), acidic (L-glutamate), or nonmetabolized (cycloleucine) acids, was examined in rats. Hepatic vagotomy enhanced both plasma glucose and glucagon concentrations after intraperitoneal injecton of alanine more than those in sham-vagotomized (control) rats, while after intraperitoneal injection of leucine, hepatic vagotomy decreased plasma glucose concentrations and enhanced plasma insulin concentrations more than in control animals. These effects, following both alanine and leucine administration, were blocked by celiac vagotomy. Glutamate, phenylalanine, and cycloleucine stimulation in hepatic-vagotomized rats caused no significant differences in plasma glucose, insulin, or glucagon levels as compared to those in sham-vagotomized rats. Celiac vagotomy alone did not affect plasma glucose, insulin, or glucagon concentrations after stimulation by these five amino acids. The physiological role of alanine and leucine sensors may be to prevent amino acid-induced exaggeraled pancreatic h e . o n e secretion and to maintain b~ood glucose homeostasis, while glutamate, phenylalanine, and cycloleucine have no effect on this paBic~c~tieneuroendocrine system.

Introduction

A number of reports have been published on vagal neural efferent pathways to the pancreatic islets modulating the secretion of insulin and glueagon [14,31], while afferent pathways which might affect this system have received little attention. Afferent neural input reaches the central nervous system from a variety of nutrient recep-

Correspondence to: Shuji Inoue, The Third Department of Internal Medicine, Yokohama City University School of Medicine, 3-9 Fukuura, Kanazawa-ku, Yokohama 236, Japan.

tots [9,21]. it has been demonstrated that receptors in the oropharyrtx stimulate preabsorptive insulin release [2,6211,26]. Further, glucoreceptc,rs in the intestine [12,13,23] and the liver [10,15,16] send information to the central nervous system via the vagus nerve, tlaereby stimulating insulin release. Amino acids are a major component of ingested foodstuffs in addition to carbohydrates, and it is now clear that preabsorptive information related to intestinal amino aci6s is an important part of the whole complex of satiety signals that govern the short-term regulation of protein intake [4,18]. Neurophysiological data reveal that

226 mesenteric nerve and vagus nerve send information to the central nervous system after amino acid infusion in the gastrointestinal tract [8,24]. Amino acids are very potent stimuli in insulin [3] and g!ucagon [1,19] secretion, however, little attention has been given to the relation between the amino acid-induced signals of hepatic origin and the pancreatic neuroendocrine system. In 1986 [27] we found that vagal arginine sensors exist in the liver and that administration of arginine into the portal vein causes a reflex inhibition of pancreatic vagus nerve activity. We also found that intraperitoneal arginine enhanced both plasma insulin and glucagon concentrations more in hepatic-vagotomized than in sham-vagotomized rats [28]. These effects were blocked by the addition of celiac vagotomy or the administration of atropine [29]. We recently found, using an electrophysiological approach, that vagal amino acid sensors exist in the liver for alanine and leucine [30]. The next step is to determine whether or not these 'amino acid sensors' besides arginine have a role in the pancreatic neuroendocrine system. The present study was designed to test the effect of hepatic vagotomy a n d / o r celiac vagotomy on neutral (L-alanine, L-leucine, and L-phenylalanine), acidic (L-glutamate), and nonmetabolized (cycloleucine) acid-induced glucose, insulin, and glucagon changes after intraperitoneal amino acid administration.

Materials and Methods

Animals 16-week-old female Sprague-Dawley rats weighing 260-280 g were housed individually in wire-bottomed staialess steel cages and exposed to 12-h light-dark cycles at constant temperature (23 +_2°C). The rats were allowed free access to laboratory chow and water except the night before the experiment.

Surgery and cannulation Three days before the experiment, surgery was performed under hexobarbital anesthesia (50 mg/kg body weight). For each experiment, four groups of eight animals each were prepared: 1) a

group of hepatic-vagotomized rats; 2) a group of hepatic-vagotomized and celiac-vagotomized rats; 3) a group of celiac-vagotomized rals; and 4) a group of sham-vagotomized (control) rats. Hepatic vagotomy and celiac vagotomy were performed as previously described [28,29]. Sham vagotomy was achieved by the same surgical technique except for the sectioning of the vagus nerve. After surgery, a catheter (silastic and polyethylene tubing (PE-50)) was inserted into the right atrium of the heart for blood samplings [22]. The open end of the catheter was exposed to the posterior neck through the subcutaneous tissue. After cannulation, the rats were returned to their cages and allowed free access to food and water.

Experimental procedures 3 days after the surgery, five experiments were performed under an unanesthetized and unrestrained state after 16 h overnight food deprivation. Aqueous amino acid solutions were prepared in concentrations of 2.5-12.5 g per 100 ml, depending upon the solubility of the amino acid in water at 37°C. Each experiment was for the purpose of determining the effect of hepatic vagotomy a n d / o r celiac vagotomy on plasma glucose, insulin, and glucagon concentrations following intraperitoneal injection of the amino acid shown: Experiment 1: L-alanine solution (0.5 g / k g body weight); Experiment 2: L-leucine suspension (0.3 g / k g body weight in a 10% gum arabic solution); Experiment 3: monosodium-L-glutamate solution (0.5 g / k g body weight); E~periment 4: L-phenylalanine solution (0.1 g/k~g body weight); and Experiment 5: cycloleue:~ne (1amino-l-cyclopentane carboxylic acid)(0.2 g / k g body weight). For glucose and hormone determinations, 0.9 ml of blood was withdrawn from the right atrium through the cardiac catheter with a heparinized syringe before (0 min) and 5, 10, 15, 30 and 60 min after the injection anJ poured into a glass tube containing 1000 U of Trasylol. Immediately after each blood sampling, the same dose of heparinized blood was replaced to minimize the influence of blood volume depletion. The blood samples were chilled in ice and plasma was separated immediately after the completion of the

227

(mg/dl) 15o1 Alanine

experiment. The plasma was stored at -80°C until the assay. At the end of each experiment, confirmation was made of hepatic a n d / o r celiac vagotomy in all animals using a binocular microscope.

A

14oI 4r 13o1

m2o1

,k--

tzO-I

/

"

--

Assay Glucose was measured in a Beckman glucose analyzer employing the glucose oxidase method. Immunoreactive insulin was measured by the modified double antibody method of Hales and Randle [5] using rat insulin as a standard (insulin assay kit, Amersham, Japan). Immunoreactive glucagon was measured by the method of hnagawa et al. [7] using a specific antiserum for pancreatic giucagon and porcine giucagon standards (Glucagon assay kit, Daiichi Radioisotope Labs., Japan).

Statist&al analys& Data were statistically analyzed by two-way (treatment x min) analysis of variance (ANOVA). When the F values proved significaut, Student's t-test was employed to identify significant differences between groups. The level of significance was P < 0.05.

Results

Experiment 1: intraperitoneal injection of L-alanine There was no significant difference in body weight change between the four experimental groups 3 days after the operations. Plasma glucose, insulin, and gluc;~.gon concentrations after intraperitoneal injection of alanine are shown in Fig. 1. Plasma glucose concentrations increased significantly after alanine stimulation in each of the four groups (basal vs. postinjected values, P < 0.05 to 0.01, respectively). In hepatic-vagotomized rats, plasma glucose concentrations resulted in a more significant elevation at 5, 10, 15 and 30 min after alanine stimulation (P < 0.05 to 0.01) than those in the other groups (Fig. 1A). Plasma insulin concentrations did not change after alanine stimulation in any group nor were there any significant differences in plasma insulin

B

(ng/ml) 2

E e,,

(pg/ml) 500 -

**~§§ ]

400I

g

=J

300-

s=

200-

vJ D

*''~

*.t,~

"*

C

'

I

~.

I

" "xT

!00O-

(I

5

II1

1'5

3()

6'0

TIME (rain) Fig. I. Time course of plasma glucose (A), insulin (B), and glueagon (C) responses to i.p. alanine. Closed circles, hepafic-vagotomized rats; closed triangles, hepatic- and celiac-vagotomized rats; open triangles, celiac-vago:omized rats; open circles, sham-vagotomized rats. Each point is ~nean + S.E., n = 8. * P < 0.05; ** P < 0.01 compared with shamvagotomized rats. t P < 0.05; .~ P < 0.01, differences between hepatic-vagotomized rats and hepatic- and celiac-vagotomized rats, § P <0.05; §§ P <0.01, differences between hepaticvagotoniized rats and celiac-vagotomized rats.

concentrations at any time among them (Fig. 1B). Plasma glucagon concentrations increased significantly after aianine stimulation in each group (basal vs. postinjeeted values; P < 0.05 to 0,01, respectively). In hepatic-vagotomized rats, plasma glucagon concentrations were more significantly increased at 5, 10, 15 and 30 rain after alanine stimulation (P < 0,05 to 0.01) than seen in the other groups (Fig. IC). There were no significant differcnces in the basal glucose, insulin or glucagon concentrations among the four groups.

228

and were significantly lower at 5, 10 and 15 min (P < 0.05 to 0.01) than those in the other groups (Fig. 2A). Plasma insulin concentrations increased significantly after ieucine stimulation in each group (basal vs. postinjected values; P < 0.05 to 0.01, respectively). In hepatic-vagotomized rats, plasma insulin concentrations were significantly higher at 5, 10, 15 and 30 min after leucine stimulation (P < 0.05 to 0.01) than those in the other groups (Fig. 2B). Plasma glucagon concentrations increased significantly after leucine stimulation in each group (basal vs. postinjected values; P < 0.05 to 0.01, respectively), however, there were no significant differences among the groups at any time (Fig. 2C). There were no significant differences in the basal glucose, insulin, or glucagon concentrations among the four groups.

A

(mg/dl)

9o

0o]

{ng/ml] 6-

*.1§

B

*1§

5 /

.~

4

\

***§~f

\

3

I"

{pg/ml) 400-

Experiment 3: intraperitoneal injection of L-glutamate

~c~ 300~

200"

E ~

100'

o" 6

~

1'O

I~ 3'o TIME (mini

e'o

Fig. 2. Time course of plasma glucose (A), insulin (B), and glucagon (C) responses to i.p. leucine. Closed circles, hepaticvagotomized rats; closed triangles, hepatic- and celiac-vagotomized rats; open triangles, celiac-vagotomized rats; open circles, sham-vagotomized rats. Each point is mean + S.E., n = 8. * P < 0.05, ** P < 0.01 compared with sham-vagotomized rats. t P < 0.05, 1: P < 0.01, differences between hepaticvagotomized rats and hepatic- and celiac-vagotomized ra,~s. § P < 0.05, §§ P < 0.01, differences between hepatic-vagotomized rats and celiac-vagotomized rats.

Experiment 2: intraperitoneai injection of L-leucine Plasma glucose, insulin, and glucagon concentrations after intraperitoneal injection of leucine are shown in Fig. 2. Plasma glucose concentrations increased significantly after leucine stimulation in sham-vagotomized rats, celiac-vagotomized rats, and hepatic- and celiac-vagotomized rats (basal vs. postinjected values; P <0.05 to 0.01, respectively). In he[atic-vagotomized rats, plasma glucose concentrations showed no statistically significant change after leucine stimulation

Plasma glucose, insulin and glucagon concentrations after intraperitoneal injection of glutamate are shown in Fig. 3. Plasma glucose and glucagon concentrations showed a statistically significant increase after glutamate stimulation (basal vs. postinjected values; P < 0.05 to 0.01, respectively), however, no changes were observed in plasma insulin concentrations after glutamate stimulation in any group. There were no significant differences in plasma glucose, insulin, or giucagon concentrations at any time among the four groups.

Experiment 4: &traperitoneal injection of L-phenylalanine Plasma glucose, insulin, and glucagon concentrations showe0 no statistically significant changes after phenylalanine stimulation in any group (data not shown), nor were there significant differences in plasma glucose, insulin, and glucagon concentrations at any time.

Experiment 5: intraperitoneal injection of cycloleuc&e Plasma glucose, insulin, and glucagon concentrations showed no statistically significant changes after cycloleucine stimulation in any group (data not shown), and there were no significant differ-

229

(mg/dl) i 20] Glutamate

!101

A

,L._.,t..

too1

..

go

_=" 8ol (ng/ml)

B 2 _ - - 2 - - - ~

OJ (pg/ml)

300".

C

20O

E loo] r.. °J o

5

I0

(5

31)

"6'0

TIME (mini

Fig. 3. Time course of plasma glucose (A), insulin (B), and glueagon (C) responses to i.p. glutamate. Closed circles, hepatic-vagotomized rats; closed triangles, hepatic- and celiacvagotomized rats; open triangles, celiac-vagotomized rats; open circles, sham-vagotomized rats. Each point is mean± S.E., n = 8 .

ences in plasma glucose, insulin, and glucagon concentrations at any time among the four groups.

Discussion We demonstrated that intraperitoneal alanine stimulation in hepatic-vagotomized rats as compared to sham-vagotomized animals is followed by marked rises in circulating glucagon levels, resulting in an elew~tion of plasma glucose, and that leucine stimulation is followed by marked rises in circulating insulin levels, resulting in reduction of plasma glucose levels. Sectioning ol the celiac vagal branches to the pancreas abolished these changes. Intraperitoneal glutamate, phenylalanine, and cycloleucine stimulation in hepatic-vagotomized rats showed no significant differences in pla:~ma glucose, insulin, or glucagon

levels as compared to those in sham-vagotomized rats. Celiac vagotomy alone did not affect plasma glucose, insulin, or glucagon concentrations after stimulation by these amino acids. These results suggested that there might be two different reflex systems from hepatic amino acid sensors, i.e. alanine and leucine sensors. The hepatic branch of the vagus nerve contains alanine- and leucine-sensitive afferent fibers which sense portal concentrations of these amino acids and send signals to the brainstem centers; this then inhibits the activities of the vagal efferent fibers to pancreatic A cells after alanine stimulation or inhibits those to pancreatic B cells after leucine stimulation. Hepatic vagotomy cuts off this inhibition and increases insulin or glucagon secretion from pancreatic islets. This explanation is based on our recent electrophysiological evidence [30] that amino acid sensors for alanine and leucine exist in the liver. The results of glutamate, phenylalanine, and cycloleucine stimulation might be explained by: 1) the hepatic branch of the vagus nerve does not contain the amino acid-sensitive afferent fibers which sense portal glutamate, phenylalanine, and cycloleucine concentrations after amino acid stimulation; or 2) the hepatic branch of this nerve contains these amino acid-sensitive afferent fibers, but due to failure of pancreatic hormone secretion by these amino acids, the amino acid sensors do not take part in the pancreatic neuroendocrine system after intraperitoneal glutamate, phenylalanine, and cycloleucine stimulation. Niijima et al. [1'7] recently reported that cephalic-phase insulin release was induced by taste stimulus of monosodium-L-glutamate, a non-specific neural excitant. Jeanningros [8] also reported that various vagal amino acid receptors including glutamate and cycloleucine are present in the small intestine. Although further studies are needed to clarif)' zhe existence of these amino acid sensors in the liver, we feel that the latter explanation may be the more plausible at present. Enhancement of pancreatic hormone secretion by hepatic vagotomy was attenuated by adding celiac vagotomy, which implies that the celiac branches of the vagus nerve consist of efferent pathways to the pancreas from these amino acid

230

'sensors'. Several studies have shown the efferent pathways and receptor mechanisms for neural metabolic receptors which affect the pancreatic neuroendocrine system. Receptors in the oropharynx are known to be important for early insulin response before a rise in glucose after food ingestion which has been abolished by vagotomy [11]. Lee and Mil!er [10] demonstrated that acute sectioning of the hepatic vagus nerve caused an increase in plasma insulin concentration and that electrical stimulation of the central end of the hepatic branch of the vagus nerve suppressed the concentration of plasma insulin compared to sham-stimulated controls. Sectioning of the celiac vagal branches to the pancreas abolished these changes [10,20]. We have found that enh.,acement of insulin and glucagon secretion by hepatic vagotomy after arginine stimulation were attenuated by adding celiac vagotomy or atropine, but not by phentolamine or propranolol [29]. These results suggest that celiac branches of the vagus nerve act as efferent pathways to the pancreas through a muscarinic receptor mechanism and sympathetic pathways are not involved in the hepatic arginine-sensor mediated pancreatic neuroendocrine system. Therefore, efferent pathways for liver amino acid sensors observed in the present study seem to be consistent with these findings of a previously reported vagal-vagal reflex mechanism. The effect of hepatic vagotomy on amino acid-induced insulin secretion seemed to have an effect opposite to that reported by Lee and Miller [10] on glucose-induced insulin secretion. This difference might be explained by electrophysiological studies which found that activity of hepatic vagal afferents was reduced after intraportal glucose [15], but increased after intraportal arginine [27], alanine and lcucine [30]. There may be another possibility: the enhancement of insulin or glucagon secretion by alanine and leucine after hepatic vagotomy could be due to modulation of the liver metabolism of amino acids by the vagotomy, and, consequently, a high concentration of the amino acids in hepaticvagotomized rats would increase insulin or glucagon secretion. This explanation is based on the fact that vagal activation affects hepatic

metabolism [25]. However, hepatic vagotomy did not alter plasma amino acid concentrations following intraperitoneal alanine and leucine, respectively (unpublished observation). Consequently, it is implausible that hepatic vagotomy modulates the amino acid metabolism in the liver. What role might the amino acid sensor system play in the physiological state? In the fasting state, these sensors do not activate any function. However, in the feeding state, excessive supply and direct stimulation of these amino acids after protein ingestion from the gastrointestinal tract may cause exaggerated insulin and,/or glucagon secretion. Under these circumstances, nerve sensors in the liver sense nutrient amino acids (arginine, alanine and leucine) and convey this information through afferent vagal nerves to the brainstem from which neural efferent pathways prevent exaggerated pancreatic hormone secretion and maintain blood glucose homeostasis. We speculate that the autonomic center probably receives a constant afferent influx of amino acid related signals through the hepatic vagus nerve, which may interact with vagal efferent mechanisms controlling centrifugal influence on appropriate pancreatic hormone secretion and provide a feedback-control of glucose homeostasis.

Acknowledgement This work was partly supported by the Nisshin Seifun Foundation.

References 1 Assam R., P,osselin, G. and Dolais, J., Effects sur la glucagon6mic des perfusions et ingestions d'acides amint~s, J. Ann. Diab,,ztol. Hftel Dieu, 7 {1967) 25-41. 2 Berthoud, H.R., Bereiter, D.A., Trimble, E.F., Siegel, E.A. and Je~mrenaud, B., Cephalic phase, reflex insulin secretion. Neuroanatomical and physiological characterization, Diabetologia, 20 ( 1981 ) 393-401. 3 Floyd, J.CJr., Fajans, S.S., Conn, J.W., Knopf, R.F. and Ruii, .i., Stimulation of insulin scctctiou by m-niftoacids, J. Clin. Invest., 45 (19661 1487-1502. 4 Geary, N., Food intake and behavioral caloric compensation after protein depletion in the rat, Physiol. Behav., 23 (1979) 1089-11(198.

231 5 Hales, C.N. and Randle, P.J., Immunoassay of insulin with insulin antibody precipitate, Biochem. J., 90 (1963) 620624. 6 Halter, J., Kulkosky, P., Woods, S., Chen, M. and Porte, D. Jr., Afferent receptors, taste perception, and pancreatic endocrine function in man, Diabetes, 24 (1975) 414. 7 lmagawa, K., Nishino, T., Shin, S., Uehara, S,, Hashimura, E., Yanaihara, C. and Yanaihara, N., Production of antiglucagon sera with a C-terminal fi'agment of pancreatic glucagon, Endocrinol. Jpn., 26 (1979) 123-131. 8 Jeanningros, R., Vagal unitary responses to intestinal amino acid infusions in the anesthetized cat: A putative signal for protein induced satiety, Physiol. Behav., 28 (1982) 9-21. 9 Lautt, W.W., Hepatic nerves: a review of their function and effects, Can. J, Physiol. Pharmacol., 58 (1980) 105-123. 10 Lee, K.C. and Miller, R.E., The hepatic vagus nerve and the neural regulation of insulin secretion, Endocrinology, ! 17 (1985) 307-314. 11 Louis-Sylvestre, J., Preabsorptive insulin release and hypoglycemia in rats, Am. J. Physiol., 230 (1976) 56-60. 12 Mei, N., Vagal glucorecegtors in the small intestine of the cat, J. Physiol. (Lond), 282 (1978) 485-506. 13 Mei, N., Arlhac, A. and Boyer, A., Nervous regulation of insulin release by the intestinal vagal glucoreceptors, J. Auton. Nerv. Syst., 4 (19'81) 351-363. 14 Miller, R.E,, Pancreatic neuroendocrinology: peripheral neural mechanisms in the regulation of the islets of Langerhans, Endocrinol. Rev., 2 (1981) 471-494. 15 Niijima, A., Afferent imlpulse discharges from glucoreceptors in the liver of the guinea pig, Ann. NY. Acad. Sci., 157 (1969) 690-700. 16 Niijima, A., Nervous re~ulatory mechanism of blood glucose levels. In Y. Katsuki et al, (Eds.), Food Intake and Chemical Senses, Japan Scientific Societies Press, Tokyo, 1977, pp, 413-426. 17 Niijima, A., Togiyama, T. and Adachi, A., Cephalic-phase insulin release induced by taste stimulus of monosodium glutamate (umami taste), Physiol. Behav., 48 (1990) 905908. 18 Novin, D. and Vanderweele, D.A., Visceral involvement in feeding: There is more to regulation than the hypothalamus. in J.M. Sprague and A.N. Epstein (Eds.), Progress in Psychobiology and Physiological Psychology, vol 7, Academic Press, New York, 1977, pp. 193-241.

19 Rocha, D.M., Faloona, G.R. and Unger, R.H., Glucagonstimulating activity of 20 amino acids in dogs, J. Cli~a. Invest., 51 (1972) 2346-2351. 20 Sakaguchi, T, and Yamaguehi, K., Effects of electrical stimulation of the hepatic vagus nerve on the plasma insulin concentration in the rat, Brain Res., 164 (1979) 314-316. 21 Sawchenko, P.E. and Friedman, M.i., Sensory functions of the liver - a review, Am. J. Physiol., 236 (Regulatory Integrative Comp. Physiol. 5) (1979) R5-R20. 22 Shah, J.H., Wongsurawat, N., Aran, P.P., Motto, G.S. and Bowser, E.N., A method for studying acute insulin secretion and glucose tolerance in unanesthetized and unrestrained rats, Diabetes, 26 (1977) 1-6, 23 Sharma, K.N., Receptor mechanisms in the alimentary tract: their excitation and functions. In C.F. Code (Ed.), Handbook of Physiology, sect. 6, chap. 17, vol. 1, American Physiological Society, Washington DC, 1967, pp. 225237. 24 Sharma, K.N, and Nasset, E.S.. Electrical activity in mesenteric nerves after perfusion of gut lumen. Am. J. Physiol,, 202 (1962) 725-730. 25 Shimazu, T., Glycogen synthetase activity in the liver: regulation by the autonomic nerves, Science, 156 (1967) 1256-1257, 26 Steffans, A.B., Influence of the oral cavity on insulin release in the rat, Am. J. Physiol., 230 (1976) 1411-1415. 27 Tanaka, K,, Indue, S., Takamura, Y., Jiang, Z.-¥. and Niijima. A, Arginine sensors in the hepato-portal system and thei~ reflex effects on pancreatic effereJJts in the rat, Neurosci Lett., 72 (1986) 69-73. 28 Tanaka, K,, Indue, S., Fujii, T. and Takamura, Y., Enhancement of insulin and glucagon secretion by arginine after hepatic vagotomy, Neurosci. Lett., 72 (1986) 74-78. 29 Tanaka. K., Indue. S., Nagase, H. and Takamura, Y., Modulati.m o[ arginine-induced insulin and glucagon secretion by the hepatic vagus nerve m the rat: effects of celiac vagotomy and administration of atropine. Endocrinology, 127 (1990) 2017-2023. 30 Tanaka, K., Indue, S., Nagase, H., Takamura, Y. and Niijima, A , Amino acid sensors sensitive to alanine and Icucine exist in the hepato-portal system in the rat, J. Auton. Nerv. Syst., 31 (1990) 41-46. 31 Woods, S.C. and Porte, D.Jr.. Neural control of the endocrine pancreas, Physiol. Rev., 54 (1974)596-619.