- Email: [email protected]
Effects of ghrelin on insulin and glucagon secretion: a study of isolated pancreatic islets and intact mice
Regulatory Peptides 118 (2004) 143 – 150 www.elsevier.com/locate/regpep
Effects of ghrelin on insulin and glucagon secretion: a study of isolated pancreatic islets and intact mice Albert Salehi *, Charlotta Dornonville de la Cour, Rolf Ha˚kanson, Ingmar Lundquist Department of Pharmacology, Institute of Physiological Sciences, University of Lund, Lund, Sweden Received 18 November 2003; received in revised form 9 December 2003; accepted 18 December 2003
Abstract We combined in vitro and in vivo methods to investigate the effects of ghrelin, a novel gastric hormone, on insulin and glucagon release. Studies of isolated mouse islets showed that ghrelin concentrations in the physiological range (0.5 – 3 nmol l 1) had no effect on glucosestimulated insulin release, while low ghrelin concentrations (1 – 100 pmol l 1) inhibited and high (0.1 and 1 Amol l 1) stimulated. The insulin response to glucose was enhanced in the presence of a high ghrelin concentration (100 nmol l 1). Glucagon release was stimulated by ghrelin (0.1 pmol l 1 to 1 Amol l 1); this effect was maintained in the presence of glucose (0 – 20 mmol l 1). In intact mice, basal plasma insulin was suppressed by 1 and 10 nmol kg 1 of ghrelin, 2 and 6 min after i.v. injection. Ghrelin (0.2 – 10 nmol kg 1 i.v.) suppressed also the glucose-stimulated insulin response and impaired the glucose tolerance (at a ghrelin dose of 3.3 nmol kg 1). Ghrelin (1 or 10 nmol kg 1 i.v.) inhibited the insulin response to the phospholipase C stimulating agent carbachol and enhanced the insulin response to the phosphodiesterase inhibitor isobutyl-methylxanthine (IBMX) but did not affect the response to the membrane-depolarizing amino acid Larginine. These observations suggest that the inhibitory effect of ghrelin on glucose-induced insulin release is in part exerted on phospholipase C pathways (and not on Ca2 +entry), while the stimulatory effect of high doses of ghrelin depends on cyclic AMP. In contrast to the spectacular glucagon-releasing effect of ghrelin in vitro, ghrelin did not raise plasma glucagon. Carbachol, IBMX and L-arginine stimulated glucagon release. These responses were impaired by ghrelin, suggesting that it suppresses the various intracellular pathways (phospholipase C, cyclic AMP and Ca2 +), that are activated by the glucagon secretagogues. Together these observations highlight (but do not explain) the different effects of ghrelin on glucagon release in vitro and in vivo. The results show that ghrelin has powerful effects on islet cells, suggesting that endogenous ghrelin may contribute to the physiological control of insulin and glucagon release. However, the narrow ‘‘window’’ of circulating ghrelin concentrations makes this doubtful. D 2004 Elsevier B.V. All rights reserved. Keywords: Islet hormone secretion; Ghrelin
1. Introduction There is a suspicion that the stomach harbours peptide hormones that contribute to the control of pancreatic hormone release. This suspicion is based on reports describing impaired insulin release following resection of the stomach both in man and experimental animals [1 – 3]. Recently, a growth hormone-releasing peptide named ghrelin was identified in extracts of the rat stomach [4] and found to be produced and stored in a population of peptide hormoneproducing cells referred to as A-like cells [5 –7]. Circulating concentrations of ghrelin are elevated by food deprivation
* Corresponding author. Tel.:+46-46222-7588; fax: +46-46222-4429. E-mail address: [email protected] (A. Salehi). 0167-0115/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2003.12.001
and lowered by food intake [6,8,9]. The signals that regulate the secretion of ghrelin from the A-like cells have not yet been identified. Whether the A-like cells and ghrelin participate in the control of islet hormone secretion remains an unresolved issue although some reports have appeared, suggesting that ghrelin either stimulates [10,11] or inhibits [12 – 14] insulin secretion. There have been claims that insulin (and/or glucose) lowers circulating ghrelin [15 – 19] (for a different view, see Ref. [20]). There are surprisingly few data on the effect of ghrelin on glucagon release; however, it was recently reported that ghrelin is without effect on glucagon secretion in the perfused rat pancreas [13]. The purpose of the present study was to examine the effects of ghrelin on insulin and glucagon release from freshly isolated pancreatic islets from mice and to further
144
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
investigate effects of ghrelin on insulin and glucagon secretion in intact, freely moving mice using different blood sampling times and different doses with or without administration of glucose and various insulin and glucagon secretagogues.
vein or intraperitoneally (5 Al g 1) at different doses, alone or in combination with glucose, carbachol, isobutyl-methylxantine (IBMX) or L-arginine (time scale and doses are specified in Results). Controls received gelatine –saline (denoted as saline). Blood was sampled as described previously [3]. 2.4. Determination of insulin and glucagon
2. Methods 2.1. Drugs and chemicals
Insulin and glucagon were determined in plasma and incubation medium by radioimmunoassay [22 –24].
Rat ghrelin-28 was a kind gift from Professors N. Yanaihara and C. Yanaihara at the Yanaihara Institute, Shizuoka, Japan. Collagenase (CLS-4) from Worthington Biochemicals (Freehold, NJ, USA) was used to prepare the pancreatic islets. Bovine serum albumin (BSA) was from ICN Biomedical (High Wycombe, UK). All other chemicals were from British Drug Houses (Poole, UK) or Merck (Darmstadt, Germany). Radioimmunoassay kits for determination of insulin were obtained from Diagnostika (Falkenberg, Sweden) and those for glucagon determination from Eurodiagnostica (Malmo¨, Sweden). The antiserum used in the glucagon assay recognizes pancreatic glucagon but not gut glucagon or gut glucagon-like peptides. 2.2. Animals Freely fed female mice of the NMRI strain (B&K, Sollentuna, Sweden), weighing 25 – 30 g (3 – 4 months old), were used. They were given a standard pellet diet (B&K) and tap water ad libitum. The experiments involved non-anaesthetized mice and were approved by the local animal welfare committee. 2.3. Experimental protocol 2.3.1. In vitro studies Pancreatic islets were isolated from mice, killed by cervical dislocation. A collagenase solution was injected into the bile –pancreatic duct, followed by excision of the pancreas and isolation of the islets by a standard digestion procedure [21]. Freshly isolated islets were preincubated for 30 min at 37 jC in Krebs– Ringer bicarbonate buffer, pH 7.4, supplemented with 10 mM Hepes, 0.1% bovine serum albumin and 1 mM glucose. Each incubation vial was gassed with 95% O2 and 5% CO2 to obtain constant pH and oxygenation. After preincubation, the buffer was changed to a medium containing different concentrations of glucose and ghrelin. The islets (10 islets in a volume of 1 ml) were incubated at 37 jC in a metabolic shaker (30 cycles per min). After 60 min of incubation, aliquots of the medium were removed for assay of insulin and glucagon. 2.3.2. In vivo studies Ghrelin was dissolved in 0.1% gelatine – 0.9% NaCl (gelatine –saline) and given as a bolus injection into a tail
Fig. 1. Dose – response effect of ghrelin on insulin and glucagon release from isolated islets. Isolated islets were incubated in a medium, containing 12 mmol l 1 glucose. Insulin (a) and glucagon (b) release was measured. Ghrelin concentrations in the medium ranged from 1 fmol l 1 to 1 Amol 1 1. Hatched area indicates the range of blood ghrelin concentrations in rats and mice during pre- and post-prandial conditions (Refs. [4,6,8, 9,13,16,34,35] and own unpublished data). Values are means F S.E.M., 10 – 12 batches of islets at each point. Each batch contained 10 islets. Significant differences versus controls (12 mmol l 1 glucose without ghrelin) are denoted by *p < 0.05, ***p < 0.001.
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
145
3. Results 3.1. Effects of ghrelin on isolated pancreatic islets 3.1.1. Insulin and glucagon secretion in response to varying concentrations of ghrelin at a fixed glucose concentration Fig. 1 illustrates the concentration –response curves for the effect of ghrelin on insulin and glucagon secretion from isolated islets in the presence of a glucose concentration (12 mmol l 1) that in itself has a modest stimulating effect on insulin secretion. Insulin release was suppressed in the presence of low concentrations of ghrelin (1 – 100 pmol l 1), while concentrations of 0.1 and 1 Amol l 1 stimulated insulin secretion. However, within the range of circulating ghrelin concentrations (0.5 nmol l 1 in freely fed mice and around 3 nmol l 1 in fasted mice (Refs. [4,6,8,9,13,16, 34,35] and own unpublished data), insulin secretion was not affected. Ghrelin stimulated glucagon secretion already at a concentration of 0.1 pmol l 1 with a near-maximal effect at approximately 1 pmol l 1 (Fig. 1).
Fig. 2. Effects of ghrelin, 100 nmol l 1, on insulin and glucagon release from isolated islets at different glucose concentrations. Isolated islets were incubated with 100 nmol l 1 ghrelin at glucose concentrations ranging from 0 to 20 mmol l 1. Insulin (a) and glucagon (b) release was measured. Values are means F S.E.M for 10 – 12 batches of islets at each point. Each batch contained 10 islets. Significant differences versus controls without ghrelin are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
2.5. Determination of plasma glucose Plasma glucose concentrations were determined enzymatically [25]. 2.6. Statistics Results were expressed as means F S.E.M. The level of significance for the difference between sets of data was assessed using Student’s unpaired t-test or analysis of variance followed by Tukey – Kramer’s test whenever appropriate. P < 0.05 was considered statistically significant.
Fig. 3. Acute effects of ghrelin on basal plasma levels of insulin, glucagon and glucose. Effect of an i.v. injection 10 nmol kg 1 of ghrelin, (n = 12) on basal plasma levels of insulin (a) and glucagon (b) at 2 and 6 min after administration. Controls were injected with gelatine + saline (saline) (n = 14).Values are means F S.E.M. Significant differences versus controls are denoted by **p < 0.01, ***p < 0.001.
146
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
3.1.2. Insulin and glucagon secretion in response to varying concentrations of glucose at a fixed ghrelin concentration Fig. 2 illustrates the effects of an insulin-mobilizing concentration of ghrelin (100 nmol l 1) on insulin and glucagon secretion in the presence of increasing glucose concentrations (0 –20 mmol l 1). The insulin response to glucose (10 – 20 mmol l 1) was enhanced in the presence of a high ghrelin concentration (100 nmol l 1) (Fig. 2a). A strong stimulatory effect of ghrelin on glucagon secretion (Fig. 2b) was evident at all glucose concentrations tested, being particularly prominent at 10– 12 mmol l 1 of glucose, which is within the physiological plasma glucose concentration range of freely fed mice [26]. 3.2. Effects of ghrelin in the intact mouse 3.2.1. Effects of low and high doses of ghrelin on basal plasma levels of insulin, glucagon and glucose 3.2.1.1. Acute effects (2 and 6 min). While 1 nmol kg 1 of ghrelin i.v. had a negligible inhibitory effect on plasma insulin (not shown), a dose of 10 nmol kg 1 suppressed it at 2 and 6 min (Fig. 3a). Plasma glucose was moderately raised by 10 nmol kg 1 at 6 min (11.5 F 0.3 vs. 10.5 F 0.3 mmol l 1; P < 0.05; n = 14) but was not affected by 1 nmol kg 1 (data not shown). The basal plasma glucagon concentration
was not affected by either 1 or 10 nmol kg 6 min after injection (Fig. 3b).
1
ghrelin 2 and
3.2.1.2. Short-term effects (15 and 40 min). We examined the short-term as opposed to acute effects of 40 nmol kg 1 of ghrelin (i.v.) on plasma insulin, glucagon and glucose. In accordance with Lee et al. [11], we found raised basal plasma insulin concentration after 15 min; basal glucagon and glucose levels were raised moderately after 15 and 40 min (data not shown). 3.2.2. Effects of ghrelin on insulin and glucagon responses to glucose 3.2.2.1. Acute effects (2 and 6 min). Injection of glucose raised the plasma insulin concentration and lowered the glucagon concentration; a dose of 3.3 mmol kg 1 i.v. of glucose is known to induce a half-maximal insulin response in mice [27]. Fig. 4 shows the effect of a wide range of ghrelin doses (from 0.1 to 40 nmol kg 1) injected i.v. approximately 15 s before the i.v. injection of 3.3 mmol kg 1 of glucose. In the low dose range (0.2 – 10 nmol kg 1), ghrelin suppressed the insulin response 2 and 6 min after the glucose challenge, while the insulin response was unaffected by ghrelin at a dose of 40 nmol kg 1. Ghrelin had negligible acute effects on the plasma glucagon levels in glucose-treated mice (Fig. 4).
Fig. 4. Effect of ghrelin on insulin and glucagon responses to an i.v. glucose load. A dose – response study of the acute effects of ghrelin (0 – 40 nmol kg 1) on plasma levels of insulin and glucagon following an i.v. glucose load (3.3 mmol kg 1). Ghrelin was given i.v. 15 s prior to i.v. glucose. Controls received gelatine – saline (saline). Blood samples were collected 2 and 6 min after the glucose injection. Values are means F S.E.M. (vertical bars). Significant differences between controls (saline + glucose) (indicated by interrupted lines) (n = 37) and ghrelin + glucose (n = 10 – 24) are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
147
3.2.2.3. Glucose tolerance. Ghrelin (3.3 nmol kg 1) was injected i.p. together with glucose (11 mmol kg 1), a dose known to induce a maximal insulin response [27]. Ghrelin impaired the glucose tolerance compared with controls receiving glucose alone (Fig. 5).
Fig. 5. Effect of ghrelin on glucose tolerance. Ghrelin, 3.3 nmol kg 1, and glucose, 11.1 mmol kg 1, were injected i.p. at time 0 (arrow) (n = 12) Controls received gelatine – saline (saline) and glucose (n = 8). Values are means F S.E.M. Significant differences versus controls are denoted by *p < 0.05.
3.2.2.2. Short-term effects (15 and 40 min). Ghrelin (40 nmol kg 1 i.v.) was given 15 s before glucose (3.3 mmol kg 1) injection (not shown). The i.v. injection of saline + glucose raised the plasma insulin concentration and lowered the glucagon concentration. Following the injection of ghrelin + glucose, the plasma insulin levels did not seem to differ from those measured in glucose-injected controls. The glucagon levels after ghrelin + glucose did not display the decrease seen after saline + glucose (not shown).
3.2.2.4. Acute effects of ghrelin on insulin and glucagon responses to L-arginine, carbachol and IBMX. Finally, we decided to test the effects of various secretagogues, known to stimulate both insulin and glucagon release through different intracellular pathways [31]. Fig. 6a shows the effect of ghrelin (10 nmol kg 1 i.v.) on the 2- and 6-min hormone responses to the membrane depolarizing agent Larginine (1.2 mmol kg 1 i.v.). Ghrelin did not affect the insulin releasing effect of L-arginine at 2 min but partially suppressed the glucagon response. Fig. 6b shows that ghrelin inhibited the hormone-releasing effects of 0.16 Amol kg 1 of the cholinergic agonist carbachol, a phospholipase C stimulator, 2 min after injection. At 6 min, however, ghrelin (1 nmol kg 1 i.v.) enhanced the cholinergic insulin response, while the glucagon response was still suppressed. Finally, we examined the effect of ghrelin on the hormone responses to the phosphodiesterase inhibitor IBMX (45 Amol kg 1 i.v.), known to stimulate both insulin and glucagon release through activation of the cyclic AMP system [28]. As shown in Fig. 6c, ghrelin (10 nmol kg 1 i.v.) enhanced the insulin releasing effect of IBMX at 2 min, while the glucagon response was inhibited at both 2 and 6 min by both 1 and 10 nmol kg 1 of ghrelin.
Fig. 6. Effect of ghrelin on insulin and glucagon responses to various secretagogues. Ghrelin (1 or 10 nmol kg 1) or gelatine – saline (saline) was given i.v. 15 s prior to i.v. administration of L-arginine (1.2 mmol kg 1) (a), carbachol (0.16 Amol kg 1) (b), or IBMX (45 Amol kg 1) (c). Basal plasma levels of insulin and glucagon are given in Fig. 3. The insulin and glucagon responses at 2 and 6 min were recorded for 6 – 13 animals in each group. Values are means F S.E.M. Significant differences versus controls are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
148
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
4. Discussion In an earlier report [3], we suggested that the stomach harbours a peptide hormone that serves to enhance the glucose-stimulated insulin response and potentiate the glucose-induced suppression of glucagon secretion. This suggestion is in line with the fact that the oxyntic mucosa of the stomach is rich in peptide hormone-producing endocrine cells, such as ECL cells and A-like cells [29,30]. The putative peptide hormone of the ECL cells is unknown, whereas ghrelin has been identified as the hormone of the A-like cells [4– 6]. The observation that circulating ghrelin levels are raised by fasting and suppressed by food intake [6,8,9] does not favour the view that ghrelin acts as an insulinotropic hormone. Nonetheless, we decided to explore the possibility that ghrelin participates in a gastroinsular axis, controlling insulin and/or glucagon secretion. Ghrelin is generally thought to be involved in the regulation of islet function. Earlier studies, however, have generated conflicting results in that ghrelin has been shown to inhibit insulin release in some experimental situations [12 – 14] and to stimulate insulin secretion in others [10,11]. These discrepancies may reflect species differences and/or differences in experimental design (different doses, different times of observation, in vitro versus in vivo experiments, etc.). In the present study of insulin and glucagon secretion in the mouse, we recorded acute (2– 6 min) as well as short-term (15 – 40 min) effects of a wide dose range of ghrelin, comparing in vitro results with those from in vivo experiments. 4.1. Insulin secretion In our studies of isolated islets, low concentrations of ghrelin (1 –100 pmol l 1) suppressed insulin release whereas concentrations of 10 nmol l 1 and higher stimulated insulin release at a modestly ‘‘hyperglycaemic’’ glucose concentration (12 mmol l 1). Whether these effects reflect a direct inhibitory effect of ghrelin on the insulin cells or whether changes in the secretion of somatostatin (or other inhibitory agents) contribute to the observed secretory pattern is unknown. Intravenous injection of ghrelin at a dose of 10 nmol kg 1 (but not at 1 nmol kg 1) lowered the basal plasma insulin concentration at 2 and 6 min. The suppression of basal plasma insulin is in line with the results of recent studies in man and in mice [12,14]. Higher doses of ghrelin were either ineffective or raised the plasma insulin concentration 15 min after injection. Administration of a low dose of ghrelin (0.1 nmol kg 1) had no effect on the insulin response to glucose while slightly higher doses (0.2 – 10 nmol kg 1) inhibited the acute insulin response to glucose. High ghrelin doses (20 and 40 nmol kg 1) were without effect. Hence, ghrelin seems to operate as a negative rather than positive modulator of the insulin response to glucose (see also Refs. [13,14]). The inhibitory effect of ghrelin on
the glucose-stimulated insulin response was reflected in a modest impairment of the glucose tolerance. Moreover, ghrelin brought about elevated basal plasma glucose levels at 6– 40 min after the injection of ghrelin. In view of the well-known fact that plasma ghrelin is elevated between meals [6,8,9], it is tempting to speculate that the post-prandial reduction in circulating ghrelin leads to a temporary disinhibition of its restraining effect on insulin release and that ghrelin may act to suppress inappropriate insulin release between meals. Conceivably, ghrelin exerts its effects directly on the insulin cells, where ghrelin receptors have been demonstrated [10]. In this context, it should be recalled that there is experimental evidence to suggest that the stomach harbours insulinotropic factors [3]. Indeed, the stomach is favourably situated to transduce food-evoked signals to the pancreatic islets. However, from the present observations it seems that ghrelin is unlikely to represent such a gastric insulinotropic signal. In accordance with the observation of Lee et al. [11], we found that an i.v. injection of a high dose of ghrelin (40 nmol kg 1) stimulated basal insulin secretion after 15 min, whereas no effect was seen after 2 or 6 min. The late onset of the insulin response might suggest that ghrelin stimulates insulin secretion in an indirect manner. From the present results however, it seems doubtful that ghrelin has a physiologically significant effect on insulin secretion, since ghrelin circulates within a quite narrow concentration range (from 0.5 nmol l 1 in freely fed animals to around 3 nmol l 1 in fasted animals (Refs. [4,6,8,9,13,16,34,35] and own unpublished data). It is a well-known fact that stimulation of insulin secretion involves several different second messenger systems. These include Ca2 +/calmodulin, which is the main messenger system for glucose-stimulated insulin release, but there is also a clear involvement of the cyclic AMP and phospholipase C systems [31]. Our in vivo experiments showed that ghrelin inhibited the insulin response to glucose (the Ca2 +/calmodulin pathway) and carbachol (the phospholipase C pathway), while it enhanced the secretagogue effect of IBMX (the cyclic AMP pathway), being without effect on the insulin response to a high dose of L-arginine (causing membrane depolarization and Ca2 +influx). It should be recalled that both glucose and L-arginine are thought to stimulate insulin secretion by elevating [Ca2 +]i. In the case of L-arginine, the effect is thought to reflect stimulated Ca2 +entry only, while in the case of glucose there is an effect not only on Ca2 +entry but also on additional mechanisms involving, for instance, phospholipase C [31]. Ghrelin markedly inhibited the insulin response to glucose but had no apparent effect on the response to L-arginine. Consequently, the inhibitory effect of ghrelin on glucosestimulated insulin release does not seem to reflect suppressed Ca2 +entry (since ghrelin did not affect the insulin response to L-arginine). Conceivably, ghrelin interferes with other mechanisms by which glucose mobilizes insulin. In
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
contrast, ghrelin was found to stimulate the insulin response to IBMX, which is known to raise cyclic AMP levels in the islets by phosphodiesterase inhibition. Hence, an activated cyclic AMP system might counteract the inhibitory effect of ghrelin exerted on the Ca2 +/calmodulin and phospholipase C systems. These observations may account for the finding that ghrelin inhibited insulin release in the low dose range and stimulated it in the high-dose range. 4.2. Glucagon secretion The effect of ghrelin on glucagon secretion has not attracted much attention. Egido et al. [13] failed to demonstrate an effect of ghrelin on arginine-stimulated glucagon release in the perfused rat pancreas. The results of our studies of isolated islets suggest that ghrelin is a potent and powerful glucagonotropic hormone in vitro. In fact, ghrelin stimulated glucagon secretion at a concentration as low as 0.1 pmol l 1. Moreover, ghrelin stimulated glucagon release regardless of the glucose concentration, although the stimulation was greatest at glucose concentrations of 10 –12 mmol l 1, i.e. within the physiological concentration range in freely fed NMRI mice [26,27]. Surprisingly, however, the effects of ghrelin on glucagon release in the intact mouse did not reproduce the spectacular in vitro effects of ghrelin. Conceivably, as yet unidentified neural and/or humoral factor(s) effectively restrain the glucagonotropic action of ghrelin in vivo. This was further underlined by our experiments with different glucagon secretagogues showing that ghrelin suppresses the glucagon response to carbachol, IBMX and L-arginine. Hence, in the in vivo situation ghrelin seems to have the ability to suppress the pathways involving phospholipase C, cyclic AMP and Ca2 +, all of which are thought to play a role in the glucagon secretory process. 4.3. Glucose metabolism At a dose of 10 nmol kg 1, ghrelin induced hyperglycaemia 6 min after its administration. When 3.3 nmol kg 1 of ghrelin was given together with glucose (11 mmol kg 1), the glucose tolerance was impaired. It has been shown previously that long-term treatment with non-peptidyl derivatives of GH-secretagogues induces hyperglycaemia in elderly subjects [32] and recently a single injection of ghrelin was found to induce acute hyperglycaemia also in young subjects [12]. The hyperglycaemic response to ghrelin could be explained partly by prompt inhibition of insulin secretion. Additional effects of ghrelin (and/or GH) on glucose handling in muscle and adipose tissue may contribute [33]. 4.4. Concluding remarks Isolated pancreatic islets responded to a low ghrelin concentration (1 –100 pmol l 1) with an increased glucagon
149
secretion and a suppressed insulin secretion, whereas high concentrations of ghrelin (100 nmol l 1 and higher) stimulated both glucagon and insulin secretion. Intravenous administration of ghrelin (0.2 – 10 nmol kg 1) was found to suppress basal as well as glucose-stimulated insulin secretion without affecting glucagon secretion. Moreover, ghrelin (1 or 10 nmol kg 1) induced an acute suppression of the glucagon response elicited through the phospholipase C, cyclic AMP and Ca2 + pathways, while the insulin response was enhanced by ghrelin activation of the cAMP system and reduced by ghrelin inhibition of the phospholipase C and Ca2 + systems. The bulk of ghrelin in the body occurs in the A-like cells of the acid-producing part of the rat stomach [6]. The question we set out to answer was whether ghrelin might be the insulinotropic hormone (gastric incretin), postulated to exist in the gastric mucosa [3]. Although ghrelin has powerful effects on the islet cells, our observations do not support this view. Most importantly, food intake suppresses ghrelin secretion [6,8,9] in contrast to what we should expect from an insulin secretagogue. From the present findings, we suggest instead that reduced circulating ghrelin concentrations, as after food intake, may raise plasma insulin and hence qualify as a negative modulator of insulin release, making the insulin cells a target for ghrelin. However, the normal blood levels of ghrelin, varying between 0.5 and 3 nmol l 1 in rats and mice (Refs. [4,6,8,9,13, 16,34,35] and own unpublished data), are such that the peptide is unlikely to contribute to the physiological control of either insulin or glucagon secretion.
Acknowledgements The study was supported by grants from the Swedish Research Council (grants 4286 and 04x-1007), the Albert Pa˚hlsson foundation, the Crafoord foundation, the Golje foundation, the Magnus Bergvall foundation, the Novo Nordisk foundation and the Medical Faculty of Lund.
References [1] Breuer RI, Moses H, Hagen TC, Zuckerman L. Gastric operations and glucose homeostasis. Gastroenterology 1972;62:1109 – 19. [2] Sudo T, Ishiyama K, Takemoto M, Kawamura M, Umemura H, Shiraha S, et al. Pancreatic endocrine function after total gastrectomy and truncal vagotomy. Am J Surg 1982;144:539 – 44. [3] Salehi A, Chen D, Ha˚kanson R, Nordin G, Lundquist I. Gastrectomy induces impaired insulin and glucagon secretion: evidence for a gastro-insular axis in mice. J Physiol 1999;514:579 – 91. [4] Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999;402:656 – 60. [5] Date Y, Kojima M, Hosoda H, Sawaguchi A, Mondal MS, Suganuma T, et al. Ghrelin, a novel growth hormone-releasing acylated peptide, is synthesized in a distinct endocrine cell type in the gastrointestinal tracts of rats and humans. Endocrinology 2000;141:4255 – 61.
150
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
[6] Dornonville de la Cour C, Bjo¨rkqvist M, Sandvik AK, Bakke I, Zhao CM, Chen D, et al. A-like cells in the rat stomach contain ghrelin and do not operate under gastrin control. Regul Pept 2001;99:141 – 50. [7] Rindi G, Necchi V, Savio A, Torsello A, Zoli M, Locatelli V, et al. Characterisation of gastric ghrelin cells in man and other mammals: studies in adult and fetal tissues. Histochem Cell Biol 2002;117: 511 – 9. [8] Tschop M, Smiley DL, Heiman ML. Ghrelin induces adiposity in rodents. Nature 2000;407:908 – 13. [9] Toshinai K, Mondal MS, Nakazato M, Date Y, Murakami N, Kojima M, et al. Upregulation of ghrelin expression in the stomach upon fasting, insulin-induced hypoglycemia, and leptin administration. Biochem Biophys Res Commun 2001;281:1220 – 5. [10] Date Y, Nakazato M, Hashiguchi S, Dezaki K, Mondal MS, Hosoda H, et al. Ghrelin is present in pancreatic alpha-cells of humans and rats and stimulates insulin secretion. Diabetes 2002;51:124 – 9. [11] Lee HM, Wang G, Englander EW, Kojima M, Greeley Jr GH. Ghrelin, a new gastrointestinal endocrine peptide that stimulates insulin secretion: enteric distribution, ontogeny, influence of endocrine, and dietary manipulations. Endocrinology 2002;143:185 – 90. [12] Broglio F, Arvat E, Benso A, Gottero C, Muccioli G, Papotti M, et al. Ghrelin, a natural GH secretagogue produced by the stomach, induces hyperglycemia and reduces insulin secretion in humans. J Clin Endocrinol Metab 2001;86:5083 – 6. [13] Egido EM, Rodriguez-Gallardo J, Silvestre RA, Marco J. Inhibitory effect of ghrelin on insulin and pancreatic somatostatin secretion. Eur J Endocrinol 2002;146:241 – 4. [14] Kvist Reimer M, Pacini G, Ahren B. Dose-dependent inhibition by ghrelin of insulin secretion in the mouse. Endocrinology 2003;144: 916 – 21. [15] Lucidi P, Murdolo G, Di Loreto C, De Cicco A, Parlanti N, Fanelli C, et al. Ghrelin is not necessary for adequate hormonal counterregulation of insulin-induced hypoglycemia. Diabetes 2002;51:2911 – 4. [16] McCowen KC, Maykel JA, Bistrian BR, Ling PR. Circulating ghrelin concentrations are lowered by intravenous glucose or hyperinsulinemic euglycemic conditions in rodents. J Endocrinol 2002;175: R7 – R11. [17] Saad MF, Bernaba B, Hwu CM, Jinagouda S, Fahmi S, Kogosov E, et al. Insulin regulates plasma ghrelin concentration. J Clin Endocrinol Metab 2002;87:3997 – 4000. [18] Shiiya T, Nakazato M, Mizuta M, Date Y, Mondal MS, Tanaka M, et al. Plasma ghrelin levels in lean and obese humans and the effect of glucose on ghrelin secretion. J Clin Endocrinol Metab 2002;87: 240 – 4. [19] Flanagan DE, Evans ML, Monsod TP, Rife F, Heptulla RA, Tamborlane WV, et al. The influence of insulin on circulating ghrelin. Am J Physiol Endocrinol Metab 2003;284:E313 – 6.
[20] Caixas A, Bashore C, Nash W, Pi-Sunyer F, Laferrere B. Insulin, unlike food intake, does not suppress ghrelin in human subjects. J Clin Endocrinol Metab 2002;87:1902. [21] Gotoh M, Maki T, Kiyoizumi T, Satomi S, Monaco AP. An improved method for isolation of mouse pancreatic islets. Transplantation 1985; 40:437 – 8. [22] Heding L. A simplified insulin radioimmunoassay method. In: Donato L, Milhaud G, Sircis J, editors. Labelled Proteins in Tracer Studies. Brussels: Euratom; 1966. p. 345 – 50. [23] Ahre´n B, Lundquist I. Glucagon immunoreactivity in plasma from normal and dystrophic mice. Diabetologia 1982;22:258 – 63. [24] Panagiotidis G, Salehi AA, Westermark P, Lundquist I. Homologous islet amyloid polypeptide: effects on plasma levels of glucagon, insulin and glucose in the mouse. Diabetes Res Clin Pract 1992;18: 167 – 71. [25] Bruss ML, Black AL. Enzymatic microdetermination of glycogen. Anal Biochem 1978;84:309 – 12. [26] Lundquist I, Panagiotidis G, Salehi A. Islet acid glucan-1,4-alphaglucosidase: a putative key enzyme in nutrient-stimulated insulin secretion. Endocrinology 1996;137:1219 – 25. [27] Lundquist I, Panagiotidis G. The relationship of islet amyloglucosidase activity and glucose-induced insulin secretion. Pancreas 1992;7: 352 – 7. ˚ kesson B, Lundquist I. Nitric oxide and hydroperoxide affect islet [28] A hormone release and Ca2 + efflux. Endocrine 1999;11:99 – 107. [29] Sundler F, Ha˚kanson R. Gastric endocrine cell typing at the light microscopic level. In: Ha˚kanson R, Sundler F, editors. The Stomach as an Endocrine Organ. Amsterdam: Elsevier; 1991. p. 9 – 26. [30] Ha˚kanson R, Chen D, Sundler F. The ECL cells. In: Johnson LR, editor. Physiology of the Gastrointestinal Tract. New York: Raven Press; 1994. p. 1171 – 84. [31] Zawalich WS, Rasmussen H. Control of insulin secretion: a model involving Ca2 +, cAMP and diacylglycerol. Mol Cell Endocrinol 1990;70:119 – 37. [32] Svensson J, Lo¨nn L, Jansson JO, Murphy G, Wyss D, Krupa D, et al. Two-month treatment of obese subjects with the oral growth hormone (GH) secretagogue MK-677 increases GH secretion, fat-free mass, and energy expenditure. J Clin Endocrinol Metab 1998;83:362 – 9. [33] Murata M, Okimura Y, Iida K, Matsumoto M, Sowa H, Kaji H, et al. Ghrelin modulates the downstream molecules of insulin signaling in hepatoma cells. J Biol Chem 2002;277:5667 – 74. [34] Beck B, Musse N, Stricker-Krongrad A. Ghrelin, macronutrient intake and dietary preferences in Long – Evans rats. Biochem Biophys Res Commun 2002;292:1031 – 5. [35] Murakami N, Hayashida T, Kuroiwa T, Nakahara K, Ida T, Mondal MS, et al. Role for central ghrelin in food intake and secretion profile of stomach ghrelin in rats. J Endocrinol 2002;174:283 – 8.
Effects of ghrelin on insulin and glucagon secretion: a study of isolated pancreatic islets and intact mice Albert Salehi *, Charlotta Dornonville de la Cour, Rolf Ha˚kanson, Ingmar Lundquist Department of Pharmacology, Institute of Physiological Sciences, University of Lund, Lund, Sweden Received 18 November 2003; received in revised form 9 December 2003; accepted 18 December 2003
Abstract We combined in vitro and in vivo methods to investigate the effects of ghrelin, a novel gastric hormone, on insulin and glucagon release. Studies of isolated mouse islets showed that ghrelin concentrations in the physiological range (0.5 – 3 nmol l 1) had no effect on glucosestimulated insulin release, while low ghrelin concentrations (1 – 100 pmol l 1) inhibited and high (0.1 and 1 Amol l 1) stimulated. The insulin response to glucose was enhanced in the presence of a high ghrelin concentration (100 nmol l 1). Glucagon release was stimulated by ghrelin (0.1 pmol l 1 to 1 Amol l 1); this effect was maintained in the presence of glucose (0 – 20 mmol l 1). In intact mice, basal plasma insulin was suppressed by 1 and 10 nmol kg 1 of ghrelin, 2 and 6 min after i.v. injection. Ghrelin (0.2 – 10 nmol kg 1 i.v.) suppressed also the glucose-stimulated insulin response and impaired the glucose tolerance (at a ghrelin dose of 3.3 nmol kg 1). Ghrelin (1 or 10 nmol kg 1 i.v.) inhibited the insulin response to the phospholipase C stimulating agent carbachol and enhanced the insulin response to the phosphodiesterase inhibitor isobutyl-methylxanthine (IBMX) but did not affect the response to the membrane-depolarizing amino acid Larginine. These observations suggest that the inhibitory effect of ghrelin on glucose-induced insulin release is in part exerted on phospholipase C pathways (and not on Ca2 +entry), while the stimulatory effect of high doses of ghrelin depends on cyclic AMP. In contrast to the spectacular glucagon-releasing effect of ghrelin in vitro, ghrelin did not raise plasma glucagon. Carbachol, IBMX and L-arginine stimulated glucagon release. These responses were impaired by ghrelin, suggesting that it suppresses the various intracellular pathways (phospholipase C, cyclic AMP and Ca2 +), that are activated by the glucagon secretagogues. Together these observations highlight (but do not explain) the different effects of ghrelin on glucagon release in vitro and in vivo. The results show that ghrelin has powerful effects on islet cells, suggesting that endogenous ghrelin may contribute to the physiological control of insulin and glucagon release. However, the narrow ‘‘window’’ of circulating ghrelin concentrations makes this doubtful. D 2004 Elsevier B.V. All rights reserved. Keywords: Islet hormone secretion; Ghrelin
1. Introduction There is a suspicion that the stomach harbours peptide hormones that contribute to the control of pancreatic hormone release. This suspicion is based on reports describing impaired insulin release following resection of the stomach both in man and experimental animals [1 – 3]. Recently, a growth hormone-releasing peptide named ghrelin was identified in extracts of the rat stomach [4] and found to be produced and stored in a population of peptide hormoneproducing cells referred to as A-like cells [5 –7]. Circulating concentrations of ghrelin are elevated by food deprivation
* Corresponding author. Tel.:+46-46222-7588; fax: +46-46222-4429. E-mail address: [email protected] (A. Salehi). 0167-0115/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2003.12.001
and lowered by food intake [6,8,9]. The signals that regulate the secretion of ghrelin from the A-like cells have not yet been identified. Whether the A-like cells and ghrelin participate in the control of islet hormone secretion remains an unresolved issue although some reports have appeared, suggesting that ghrelin either stimulates [10,11] or inhibits [12 – 14] insulin secretion. There have been claims that insulin (and/or glucose) lowers circulating ghrelin [15 – 19] (for a different view, see Ref. [20]). There are surprisingly few data on the effect of ghrelin on glucagon release; however, it was recently reported that ghrelin is without effect on glucagon secretion in the perfused rat pancreas [13]. The purpose of the present study was to examine the effects of ghrelin on insulin and glucagon release from freshly isolated pancreatic islets from mice and to further
144
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
investigate effects of ghrelin on insulin and glucagon secretion in intact, freely moving mice using different blood sampling times and different doses with or without administration of glucose and various insulin and glucagon secretagogues.
vein or intraperitoneally (5 Al g 1) at different doses, alone or in combination with glucose, carbachol, isobutyl-methylxantine (IBMX) or L-arginine (time scale and doses are specified in Results). Controls received gelatine –saline (denoted as saline). Blood was sampled as described previously [3]. 2.4. Determination of insulin and glucagon
2. Methods 2.1. Drugs and chemicals
Insulin and glucagon were determined in plasma and incubation medium by radioimmunoassay [22 –24].
Rat ghrelin-28 was a kind gift from Professors N. Yanaihara and C. Yanaihara at the Yanaihara Institute, Shizuoka, Japan. Collagenase (CLS-4) from Worthington Biochemicals (Freehold, NJ, USA) was used to prepare the pancreatic islets. Bovine serum albumin (BSA) was from ICN Biomedical (High Wycombe, UK). All other chemicals were from British Drug Houses (Poole, UK) or Merck (Darmstadt, Germany). Radioimmunoassay kits for determination of insulin were obtained from Diagnostika (Falkenberg, Sweden) and those for glucagon determination from Eurodiagnostica (Malmo¨, Sweden). The antiserum used in the glucagon assay recognizes pancreatic glucagon but not gut glucagon or gut glucagon-like peptides. 2.2. Animals Freely fed female mice of the NMRI strain (B&K, Sollentuna, Sweden), weighing 25 – 30 g (3 – 4 months old), were used. They were given a standard pellet diet (B&K) and tap water ad libitum. The experiments involved non-anaesthetized mice and were approved by the local animal welfare committee. 2.3. Experimental protocol 2.3.1. In vitro studies Pancreatic islets were isolated from mice, killed by cervical dislocation. A collagenase solution was injected into the bile –pancreatic duct, followed by excision of the pancreas and isolation of the islets by a standard digestion procedure [21]. Freshly isolated islets were preincubated for 30 min at 37 jC in Krebs– Ringer bicarbonate buffer, pH 7.4, supplemented with 10 mM Hepes, 0.1% bovine serum albumin and 1 mM glucose. Each incubation vial was gassed with 95% O2 and 5% CO2 to obtain constant pH and oxygenation. After preincubation, the buffer was changed to a medium containing different concentrations of glucose and ghrelin. The islets (10 islets in a volume of 1 ml) were incubated at 37 jC in a metabolic shaker (30 cycles per min). After 60 min of incubation, aliquots of the medium were removed for assay of insulin and glucagon. 2.3.2. In vivo studies Ghrelin was dissolved in 0.1% gelatine – 0.9% NaCl (gelatine –saline) and given as a bolus injection into a tail
Fig. 1. Dose – response effect of ghrelin on insulin and glucagon release from isolated islets. Isolated islets were incubated in a medium, containing 12 mmol l 1 glucose. Insulin (a) and glucagon (b) release was measured. Ghrelin concentrations in the medium ranged from 1 fmol l 1 to 1 Amol 1 1. Hatched area indicates the range of blood ghrelin concentrations in rats and mice during pre- and post-prandial conditions (Refs. [4,6,8, 9,13,16,34,35] and own unpublished data). Values are means F S.E.M., 10 – 12 batches of islets at each point. Each batch contained 10 islets. Significant differences versus controls (12 mmol l 1 glucose without ghrelin) are denoted by *p < 0.05, ***p < 0.001.
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
145
3. Results 3.1. Effects of ghrelin on isolated pancreatic islets 3.1.1. Insulin and glucagon secretion in response to varying concentrations of ghrelin at a fixed glucose concentration Fig. 1 illustrates the concentration –response curves for the effect of ghrelin on insulin and glucagon secretion from isolated islets in the presence of a glucose concentration (12 mmol l 1) that in itself has a modest stimulating effect on insulin secretion. Insulin release was suppressed in the presence of low concentrations of ghrelin (1 – 100 pmol l 1), while concentrations of 0.1 and 1 Amol l 1 stimulated insulin secretion. However, within the range of circulating ghrelin concentrations (0.5 nmol l 1 in freely fed mice and around 3 nmol l 1 in fasted mice (Refs. [4,6,8,9,13,16, 34,35] and own unpublished data), insulin secretion was not affected. Ghrelin stimulated glucagon secretion already at a concentration of 0.1 pmol l 1 with a near-maximal effect at approximately 1 pmol l 1 (Fig. 1).
Fig. 2. Effects of ghrelin, 100 nmol l 1, on insulin and glucagon release from isolated islets at different glucose concentrations. Isolated islets were incubated with 100 nmol l 1 ghrelin at glucose concentrations ranging from 0 to 20 mmol l 1. Insulin (a) and glucagon (b) release was measured. Values are means F S.E.M for 10 – 12 batches of islets at each point. Each batch contained 10 islets. Significant differences versus controls without ghrelin are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
2.5. Determination of plasma glucose Plasma glucose concentrations were determined enzymatically [25]. 2.6. Statistics Results were expressed as means F S.E.M. The level of significance for the difference between sets of data was assessed using Student’s unpaired t-test or analysis of variance followed by Tukey – Kramer’s test whenever appropriate. P < 0.05 was considered statistically significant.
Fig. 3. Acute effects of ghrelin on basal plasma levels of insulin, glucagon and glucose. Effect of an i.v. injection 10 nmol kg 1 of ghrelin, (n = 12) on basal plasma levels of insulin (a) and glucagon (b) at 2 and 6 min after administration. Controls were injected with gelatine + saline (saline) (n = 14).Values are means F S.E.M. Significant differences versus controls are denoted by **p < 0.01, ***p < 0.001.
146
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
3.1.2. Insulin and glucagon secretion in response to varying concentrations of glucose at a fixed ghrelin concentration Fig. 2 illustrates the effects of an insulin-mobilizing concentration of ghrelin (100 nmol l 1) on insulin and glucagon secretion in the presence of increasing glucose concentrations (0 –20 mmol l 1). The insulin response to glucose (10 – 20 mmol l 1) was enhanced in the presence of a high ghrelin concentration (100 nmol l 1) (Fig. 2a). A strong stimulatory effect of ghrelin on glucagon secretion (Fig. 2b) was evident at all glucose concentrations tested, being particularly prominent at 10– 12 mmol l 1 of glucose, which is within the physiological plasma glucose concentration range of freely fed mice [26]. 3.2. Effects of ghrelin in the intact mouse 3.2.1. Effects of low and high doses of ghrelin on basal plasma levels of insulin, glucagon and glucose 3.2.1.1. Acute effects (2 and 6 min). While 1 nmol kg 1 of ghrelin i.v. had a negligible inhibitory effect on plasma insulin (not shown), a dose of 10 nmol kg 1 suppressed it at 2 and 6 min (Fig. 3a). Plasma glucose was moderately raised by 10 nmol kg 1 at 6 min (11.5 F 0.3 vs. 10.5 F 0.3 mmol l 1; P < 0.05; n = 14) but was not affected by 1 nmol kg 1 (data not shown). The basal plasma glucagon concentration
was not affected by either 1 or 10 nmol kg 6 min after injection (Fig. 3b).
1
ghrelin 2 and
3.2.1.2. Short-term effects (15 and 40 min). We examined the short-term as opposed to acute effects of 40 nmol kg 1 of ghrelin (i.v.) on plasma insulin, glucagon and glucose. In accordance with Lee et al. [11], we found raised basal plasma insulin concentration after 15 min; basal glucagon and glucose levels were raised moderately after 15 and 40 min (data not shown). 3.2.2. Effects of ghrelin on insulin and glucagon responses to glucose 3.2.2.1. Acute effects (2 and 6 min). Injection of glucose raised the plasma insulin concentration and lowered the glucagon concentration; a dose of 3.3 mmol kg 1 i.v. of glucose is known to induce a half-maximal insulin response in mice [27]. Fig. 4 shows the effect of a wide range of ghrelin doses (from 0.1 to 40 nmol kg 1) injected i.v. approximately 15 s before the i.v. injection of 3.3 mmol kg 1 of glucose. In the low dose range (0.2 – 10 nmol kg 1), ghrelin suppressed the insulin response 2 and 6 min after the glucose challenge, while the insulin response was unaffected by ghrelin at a dose of 40 nmol kg 1. Ghrelin had negligible acute effects on the plasma glucagon levels in glucose-treated mice (Fig. 4).
Fig. 4. Effect of ghrelin on insulin and glucagon responses to an i.v. glucose load. A dose – response study of the acute effects of ghrelin (0 – 40 nmol kg 1) on plasma levels of insulin and glucagon following an i.v. glucose load (3.3 mmol kg 1). Ghrelin was given i.v. 15 s prior to i.v. glucose. Controls received gelatine – saline (saline). Blood samples were collected 2 and 6 min after the glucose injection. Values are means F S.E.M. (vertical bars). Significant differences between controls (saline + glucose) (indicated by interrupted lines) (n = 37) and ghrelin + glucose (n = 10 – 24) are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
147
3.2.2.3. Glucose tolerance. Ghrelin (3.3 nmol kg 1) was injected i.p. together with glucose (11 mmol kg 1), a dose known to induce a maximal insulin response [27]. Ghrelin impaired the glucose tolerance compared with controls receiving glucose alone (Fig. 5).
Fig. 5. Effect of ghrelin on glucose tolerance. Ghrelin, 3.3 nmol kg 1, and glucose, 11.1 mmol kg 1, were injected i.p. at time 0 (arrow) (n = 12) Controls received gelatine – saline (saline) and glucose (n = 8). Values are means F S.E.M. Significant differences versus controls are denoted by *p < 0.05.
3.2.2.2. Short-term effects (15 and 40 min). Ghrelin (40 nmol kg 1 i.v.) was given 15 s before glucose (3.3 mmol kg 1) injection (not shown). The i.v. injection of saline + glucose raised the plasma insulin concentration and lowered the glucagon concentration. Following the injection of ghrelin + glucose, the plasma insulin levels did not seem to differ from those measured in glucose-injected controls. The glucagon levels after ghrelin + glucose did not display the decrease seen after saline + glucose (not shown).
3.2.2.4. Acute effects of ghrelin on insulin and glucagon responses to L-arginine, carbachol and IBMX. Finally, we decided to test the effects of various secretagogues, known to stimulate both insulin and glucagon release through different intracellular pathways [31]. Fig. 6a shows the effect of ghrelin (10 nmol kg 1 i.v.) on the 2- and 6-min hormone responses to the membrane depolarizing agent Larginine (1.2 mmol kg 1 i.v.). Ghrelin did not affect the insulin releasing effect of L-arginine at 2 min but partially suppressed the glucagon response. Fig. 6b shows that ghrelin inhibited the hormone-releasing effects of 0.16 Amol kg 1 of the cholinergic agonist carbachol, a phospholipase C stimulator, 2 min after injection. At 6 min, however, ghrelin (1 nmol kg 1 i.v.) enhanced the cholinergic insulin response, while the glucagon response was still suppressed. Finally, we examined the effect of ghrelin on the hormone responses to the phosphodiesterase inhibitor IBMX (45 Amol kg 1 i.v.), known to stimulate both insulin and glucagon release through activation of the cyclic AMP system [28]. As shown in Fig. 6c, ghrelin (10 nmol kg 1 i.v.) enhanced the insulin releasing effect of IBMX at 2 min, while the glucagon response was inhibited at both 2 and 6 min by both 1 and 10 nmol kg 1 of ghrelin.
Fig. 6. Effect of ghrelin on insulin and glucagon responses to various secretagogues. Ghrelin (1 or 10 nmol kg 1) or gelatine – saline (saline) was given i.v. 15 s prior to i.v. administration of L-arginine (1.2 mmol kg 1) (a), carbachol (0.16 Amol kg 1) (b), or IBMX (45 Amol kg 1) (c). Basal plasma levels of insulin and glucagon are given in Fig. 3. The insulin and glucagon responses at 2 and 6 min were recorded for 6 – 13 animals in each group. Values are means F S.E.M. Significant differences versus controls are denoted by *p < 0.05, **p < 0.01, ***p < 0.001.
148
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
4. Discussion In an earlier report [3], we suggested that the stomach harbours a peptide hormone that serves to enhance the glucose-stimulated insulin response and potentiate the glucose-induced suppression of glucagon secretion. This suggestion is in line with the fact that the oxyntic mucosa of the stomach is rich in peptide hormone-producing endocrine cells, such as ECL cells and A-like cells [29,30]. The putative peptide hormone of the ECL cells is unknown, whereas ghrelin has been identified as the hormone of the A-like cells [4– 6]. The observation that circulating ghrelin levels are raised by fasting and suppressed by food intake [6,8,9] does not favour the view that ghrelin acts as an insulinotropic hormone. Nonetheless, we decided to explore the possibility that ghrelin participates in a gastroinsular axis, controlling insulin and/or glucagon secretion. Ghrelin is generally thought to be involved in the regulation of islet function. Earlier studies, however, have generated conflicting results in that ghrelin has been shown to inhibit insulin release in some experimental situations [12 – 14] and to stimulate insulin secretion in others [10,11]. These discrepancies may reflect species differences and/or differences in experimental design (different doses, different times of observation, in vitro versus in vivo experiments, etc.). In the present study of insulin and glucagon secretion in the mouse, we recorded acute (2– 6 min) as well as short-term (15 – 40 min) effects of a wide dose range of ghrelin, comparing in vitro results with those from in vivo experiments. 4.1. Insulin secretion In our studies of isolated islets, low concentrations of ghrelin (1 –100 pmol l 1) suppressed insulin release whereas concentrations of 10 nmol l 1 and higher stimulated insulin release at a modestly ‘‘hyperglycaemic’’ glucose concentration (12 mmol l 1). Whether these effects reflect a direct inhibitory effect of ghrelin on the insulin cells or whether changes in the secretion of somatostatin (or other inhibitory agents) contribute to the observed secretory pattern is unknown. Intravenous injection of ghrelin at a dose of 10 nmol kg 1 (but not at 1 nmol kg 1) lowered the basal plasma insulin concentration at 2 and 6 min. The suppression of basal plasma insulin is in line with the results of recent studies in man and in mice [12,14]. Higher doses of ghrelin were either ineffective or raised the plasma insulin concentration 15 min after injection. Administration of a low dose of ghrelin (0.1 nmol kg 1) had no effect on the insulin response to glucose while slightly higher doses (0.2 – 10 nmol kg 1) inhibited the acute insulin response to glucose. High ghrelin doses (20 and 40 nmol kg 1) were without effect. Hence, ghrelin seems to operate as a negative rather than positive modulator of the insulin response to glucose (see also Refs. [13,14]). The inhibitory effect of ghrelin on
the glucose-stimulated insulin response was reflected in a modest impairment of the glucose tolerance. Moreover, ghrelin brought about elevated basal plasma glucose levels at 6– 40 min after the injection of ghrelin. In view of the well-known fact that plasma ghrelin is elevated between meals [6,8,9], it is tempting to speculate that the post-prandial reduction in circulating ghrelin leads to a temporary disinhibition of its restraining effect on insulin release and that ghrelin may act to suppress inappropriate insulin release between meals. Conceivably, ghrelin exerts its effects directly on the insulin cells, where ghrelin receptors have been demonstrated [10]. In this context, it should be recalled that there is experimental evidence to suggest that the stomach harbours insulinotropic factors [3]. Indeed, the stomach is favourably situated to transduce food-evoked signals to the pancreatic islets. However, from the present observations it seems that ghrelin is unlikely to represent such a gastric insulinotropic signal. In accordance with the observation of Lee et al. [11], we found that an i.v. injection of a high dose of ghrelin (40 nmol kg 1) stimulated basal insulin secretion after 15 min, whereas no effect was seen after 2 or 6 min. The late onset of the insulin response might suggest that ghrelin stimulates insulin secretion in an indirect manner. From the present results however, it seems doubtful that ghrelin has a physiologically significant effect on insulin secretion, since ghrelin circulates within a quite narrow concentration range (from 0.5 nmol l 1 in freely fed animals to around 3 nmol l 1 in fasted animals (Refs. [4,6,8,9,13,16,34,35] and own unpublished data). It is a well-known fact that stimulation of insulin secretion involves several different second messenger systems. These include Ca2 +/calmodulin, which is the main messenger system for glucose-stimulated insulin release, but there is also a clear involvement of the cyclic AMP and phospholipase C systems [31]. Our in vivo experiments showed that ghrelin inhibited the insulin response to glucose (the Ca2 +/calmodulin pathway) and carbachol (the phospholipase C pathway), while it enhanced the secretagogue effect of IBMX (the cyclic AMP pathway), being without effect on the insulin response to a high dose of L-arginine (causing membrane depolarization and Ca2 +influx). It should be recalled that both glucose and L-arginine are thought to stimulate insulin secretion by elevating [Ca2 +]i. In the case of L-arginine, the effect is thought to reflect stimulated Ca2 +entry only, while in the case of glucose there is an effect not only on Ca2 +entry but also on additional mechanisms involving, for instance, phospholipase C [31]. Ghrelin markedly inhibited the insulin response to glucose but had no apparent effect on the response to L-arginine. Consequently, the inhibitory effect of ghrelin on glucosestimulated insulin release does not seem to reflect suppressed Ca2 +entry (since ghrelin did not affect the insulin response to L-arginine). Conceivably, ghrelin interferes with other mechanisms by which glucose mobilizes insulin. In
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
contrast, ghrelin was found to stimulate the insulin response to IBMX, which is known to raise cyclic AMP levels in the islets by phosphodiesterase inhibition. Hence, an activated cyclic AMP system might counteract the inhibitory effect of ghrelin exerted on the Ca2 +/calmodulin and phospholipase C systems. These observations may account for the finding that ghrelin inhibited insulin release in the low dose range and stimulated it in the high-dose range. 4.2. Glucagon secretion The effect of ghrelin on glucagon secretion has not attracted much attention. Egido et al. [13] failed to demonstrate an effect of ghrelin on arginine-stimulated glucagon release in the perfused rat pancreas. The results of our studies of isolated islets suggest that ghrelin is a potent and powerful glucagonotropic hormone in vitro. In fact, ghrelin stimulated glucagon secretion at a concentration as low as 0.1 pmol l 1. Moreover, ghrelin stimulated glucagon release regardless of the glucose concentration, although the stimulation was greatest at glucose concentrations of 10 –12 mmol l 1, i.e. within the physiological concentration range in freely fed NMRI mice [26,27]. Surprisingly, however, the effects of ghrelin on glucagon release in the intact mouse did not reproduce the spectacular in vitro effects of ghrelin. Conceivably, as yet unidentified neural and/or humoral factor(s) effectively restrain the glucagonotropic action of ghrelin in vivo. This was further underlined by our experiments with different glucagon secretagogues showing that ghrelin suppresses the glucagon response to carbachol, IBMX and L-arginine. Hence, in the in vivo situation ghrelin seems to have the ability to suppress the pathways involving phospholipase C, cyclic AMP and Ca2 +, all of which are thought to play a role in the glucagon secretory process. 4.3. Glucose metabolism At a dose of 10 nmol kg 1, ghrelin induced hyperglycaemia 6 min after its administration. When 3.3 nmol kg 1 of ghrelin was given together with glucose (11 mmol kg 1), the glucose tolerance was impaired. It has been shown previously that long-term treatment with non-peptidyl derivatives of GH-secretagogues induces hyperglycaemia in elderly subjects [32] and recently a single injection of ghrelin was found to induce acute hyperglycaemia also in young subjects [12]. The hyperglycaemic response to ghrelin could be explained partly by prompt inhibition of insulin secretion. Additional effects of ghrelin (and/or GH) on glucose handling in muscle and adipose tissue may contribute [33]. 4.4. Concluding remarks Isolated pancreatic islets responded to a low ghrelin concentration (1 –100 pmol l 1) with an increased glucagon
149
secretion and a suppressed insulin secretion, whereas high concentrations of ghrelin (100 nmol l 1 and higher) stimulated both glucagon and insulin secretion. Intravenous administration of ghrelin (0.2 – 10 nmol kg 1) was found to suppress basal as well as glucose-stimulated insulin secretion without affecting glucagon secretion. Moreover, ghrelin (1 or 10 nmol kg 1) induced an acute suppression of the glucagon response elicited through the phospholipase C, cyclic AMP and Ca2 + pathways, while the insulin response was enhanced by ghrelin activation of the cAMP system and reduced by ghrelin inhibition of the phospholipase C and Ca2 + systems. The bulk of ghrelin in the body occurs in the A-like cells of the acid-producing part of the rat stomach [6]. The question we set out to answer was whether ghrelin might be the insulinotropic hormone (gastric incretin), postulated to exist in the gastric mucosa [3]. Although ghrelin has powerful effects on the islet cells, our observations do not support this view. Most importantly, food intake suppresses ghrelin secretion [6,8,9] in contrast to what we should expect from an insulin secretagogue. From the present findings, we suggest instead that reduced circulating ghrelin concentrations, as after food intake, may raise plasma insulin and hence qualify as a negative modulator of insulin release, making the insulin cells a target for ghrelin. However, the normal blood levels of ghrelin, varying between 0.5 and 3 nmol l 1 in rats and mice (Refs. [4,6,8,9,13, 16,34,35] and own unpublished data), are such that the peptide is unlikely to contribute to the physiological control of either insulin or glucagon secretion.
Acknowledgements The study was supported by grants from the Swedish Research Council (grants 4286 and 04x-1007), the Albert Pa˚hlsson foundation, the Crafoord foundation, the Golje foundation, the Magnus Bergvall foundation, the Novo Nordisk foundation and the Medical Faculty of Lund.
References [1] Breuer RI, Moses H, Hagen TC, Zuckerman L. Gastric operations and glucose homeostasis. Gastroenterology 1972;62:1109 – 19. [2] Sudo T, Ishiyama K, Takemoto M, Kawamura M, Umemura H, Shiraha S, et al. Pancreatic endocrine function after total gastrectomy and truncal vagotomy. Am J Surg 1982;144:539 – 44. [3] Salehi A, Chen D, Ha˚kanson R, Nordin G, Lundquist I. Gastrectomy induces impaired insulin and glucagon secretion: evidence for a gastro-insular axis in mice. J Physiol 1999;514:579 – 91. [4] Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999;402:656 – 60. [5] Date Y, Kojima M, Hosoda H, Sawaguchi A, Mondal MS, Suganuma T, et al. Ghrelin, a novel growth hormone-releasing acylated peptide, is synthesized in a distinct endocrine cell type in the gastrointestinal tracts of rats and humans. Endocrinology 2000;141:4255 – 61.
150
A. Salehi et al. / Regulatory Peptides 118 (2004) 143–150
[6] Dornonville de la Cour C, Bjo¨rkqvist M, Sandvik AK, Bakke I, Zhao CM, Chen D, et al. A-like cells in the rat stomach contain ghrelin and do not operate under gastrin control. Regul Pept 2001;99:141 – 50. [7] Rindi G, Necchi V, Savio A, Torsello A, Zoli M, Locatelli V, et al. Characterisation of gastric ghrelin cells in man and other mammals: studies in adult and fetal tissues. Histochem Cell Biol 2002;117: 511 – 9. [8] Tschop M, Smiley DL, Heiman ML. Ghrelin induces adiposity in rodents. Nature 2000;407:908 – 13. [9] Toshinai K, Mondal MS, Nakazato M, Date Y, Murakami N, Kojima M, et al. Upregulation of ghrelin expression in the stomach upon fasting, insulin-induced hypoglycemia, and leptin administration. Biochem Biophys Res Commun 2001;281:1220 – 5. [10] Date Y, Nakazato M, Hashiguchi S, Dezaki K, Mondal MS, Hosoda H, et al. Ghrelin is present in pancreatic alpha-cells of humans and rats and stimulates insulin secretion. Diabetes 2002;51:124 – 9. [11] Lee HM, Wang G, Englander EW, Kojima M, Greeley Jr GH. Ghrelin, a new gastrointestinal endocrine peptide that stimulates insulin secretion: enteric distribution, ontogeny, influence of endocrine, and dietary manipulations. Endocrinology 2002;143:185 – 90. [12] Broglio F, Arvat E, Benso A, Gottero C, Muccioli G, Papotti M, et al. Ghrelin, a natural GH secretagogue produced by the stomach, induces hyperglycemia and reduces insulin secretion in humans. J Clin Endocrinol Metab 2001;86:5083 – 6. [13] Egido EM, Rodriguez-Gallardo J, Silvestre RA, Marco J. Inhibitory effect of ghrelin on insulin and pancreatic somatostatin secretion. Eur J Endocrinol 2002;146:241 – 4. [14] Kvist Reimer M, Pacini G, Ahren B. Dose-dependent inhibition by ghrelin of insulin secretion in the mouse. Endocrinology 2003;144: 916 – 21. [15] Lucidi P, Murdolo G, Di Loreto C, De Cicco A, Parlanti N, Fanelli C, et al. Ghrelin is not necessary for adequate hormonal counterregulation of insulin-induced hypoglycemia. Diabetes 2002;51:2911 – 4. [16] McCowen KC, Maykel JA, Bistrian BR, Ling PR. Circulating ghrelin concentrations are lowered by intravenous glucose or hyperinsulinemic euglycemic conditions in rodents. J Endocrinol 2002;175: R7 – R11. [17] Saad MF, Bernaba B, Hwu CM, Jinagouda S, Fahmi S, Kogosov E, et al. Insulin regulates plasma ghrelin concentration. J Clin Endocrinol Metab 2002;87:3997 – 4000. [18] Shiiya T, Nakazato M, Mizuta M, Date Y, Mondal MS, Tanaka M, et al. Plasma ghrelin levels in lean and obese humans and the effect of glucose on ghrelin secretion. J Clin Endocrinol Metab 2002;87: 240 – 4. [19] Flanagan DE, Evans ML, Monsod TP, Rife F, Heptulla RA, Tamborlane WV, et al. The influence of insulin on circulating ghrelin. Am J Physiol Endocrinol Metab 2003;284:E313 – 6.
[20] Caixas A, Bashore C, Nash W, Pi-Sunyer F, Laferrere B. Insulin, unlike food intake, does not suppress ghrelin in human subjects. J Clin Endocrinol Metab 2002;87:1902. [21] Gotoh M, Maki T, Kiyoizumi T, Satomi S, Monaco AP. An improved method for isolation of mouse pancreatic islets. Transplantation 1985; 40:437 – 8. [22] Heding L. A simplified insulin radioimmunoassay method. In: Donato L, Milhaud G, Sircis J, editors. Labelled Proteins in Tracer Studies. Brussels: Euratom; 1966. p. 345 – 50. [23] Ahre´n B, Lundquist I. Glucagon immunoreactivity in plasma from normal and dystrophic mice. Diabetologia 1982;22:258 – 63. [24] Panagiotidis G, Salehi AA, Westermark P, Lundquist I. Homologous islet amyloid polypeptide: effects on plasma levels of glucagon, insulin and glucose in the mouse. Diabetes Res Clin Pract 1992;18: 167 – 71. [25] Bruss ML, Black AL. Enzymatic microdetermination of glycogen. Anal Biochem 1978;84:309 – 12. [26] Lundquist I, Panagiotidis G, Salehi A. Islet acid glucan-1,4-alphaglucosidase: a putative key enzyme in nutrient-stimulated insulin secretion. Endocrinology 1996;137:1219 – 25. [27] Lundquist I, Panagiotidis G. The relationship of islet amyloglucosidase activity and glucose-induced insulin secretion. Pancreas 1992;7: 352 – 7. ˚ kesson B, Lundquist I. Nitric oxide and hydroperoxide affect islet [28] A hormone release and Ca2 + efflux. Endocrine 1999;11:99 – 107. [29] Sundler F, Ha˚kanson R. Gastric endocrine cell typing at the light microscopic level. In: Ha˚kanson R, Sundler F, editors. The Stomach as an Endocrine Organ. Amsterdam: Elsevier; 1991. p. 9 – 26. [30] Ha˚kanson R, Chen D, Sundler F. The ECL cells. In: Johnson LR, editor. Physiology of the Gastrointestinal Tract. New York: Raven Press; 1994. p. 1171 – 84. [31] Zawalich WS, Rasmussen H. Control of insulin secretion: a model involving Ca2 +, cAMP and diacylglycerol. Mol Cell Endocrinol 1990;70:119 – 37. [32] Svensson J, Lo¨nn L, Jansson JO, Murphy G, Wyss D, Krupa D, et al. Two-month treatment of obese subjects with the oral growth hormone (GH) secretagogue MK-677 increases GH secretion, fat-free mass, and energy expenditure. J Clin Endocrinol Metab 1998;83:362 – 9. [33] Murata M, Okimura Y, Iida K, Matsumoto M, Sowa H, Kaji H, et al. Ghrelin modulates the downstream molecules of insulin signaling in hepatoma cells. J Biol Chem 2002;277:5667 – 74. [34] Beck B, Musse N, Stricker-Krongrad A. Ghrelin, macronutrient intake and dietary preferences in Long – Evans rats. Biochem Biophys Res Commun 2002;292:1031 – 5. [35] Murakami N, Hayashida T, Kuroiwa T, Nakahara K, Ida T, Mondal MS, et al. Role for central ghrelin in food intake and secretion profile of stomach ghrelin in rats. J Endocrinol 2002;174:283 – 8.