Diabetologia DOI 10.1007/s00125-014-3221-0
REVIEW
Insulin release: the receptor hypothesis Willy J. Malaisse
Received: 16 January 2014 / Accepted: 28 February 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract It is currently believed that the stimulation of insulin release by nutrient secretagogues reflects their capacity to act as fuel in pancreatic islet beta cells. In this review, it is proposed that such a fuel concept is not incompatible with a receptor hypothesis postulating the participation of cellsurface receptors in the recognition of selected nutrients as insulinotropic agents. Pursuant to this, attention is drawn to such matters as the anomeric specificity of the beta cell secretory response to D-glucose and its perturbation in diabetes mellitus, the insulinotropic action of artificial sweeteners, the possible role of bitter taste receptors in the stimulation of insulin secretion by L-glucose pentaacetate, the recently documented presence of cell-surface sweet taste receptors in insulinproducing cells, the multimodal signalling process resulting from the activation of these latter receptors, and the presence in beta cells of a sweet taste receptor mediating the fructoseinduced potentiation of glucose-stimulated insulin secretion. Keywords Bitter taste receptor . Insulin release . Pancreatic islet beta cells . Sweet taste receptor In an article entitled ‘Insulin release: the fuel hypothesis’ published in 1979, it was proposed that the capacity of nutrients to act as fuel in pancreatic islet cells may account for their effects on the secretion of insulin and possibly other pancreatic hormones [1]. This fuel concept was mainly based on such findings as the parallels between the metabolic fate and insulinotropic action of several carbohydrates, including the α- and β-anomers of D-glucose or D-mannose, and the stimulation of insulin evoked either in the absence of extracellular D-glucose by glycogenolysis and glycolysis from endogenous stores of W. J. Malaisse (*) Department of Biochemistry, Université Libre de Bruxelles, 808 Route de Lennik, 1070 Brussels, Belgium e-mail: [email protected]
glycogen, by the poorly oxidised analogue of pyruvate, 3-phenylpyruvate, or by selected keto and amino acids, including the analogue of L-leucine, 2-amino-bicyclo[2,2,1]heptane-2carboxylic acid [1, 2]. The fuel concept obviously differs from the view that nutrient secretagogues are recognised by islet cells through the intervention of stereospecific receptors, possibly located at the cell membrane [3]. The main aim of this review is to illustrate the pertinence of the receptor hypothesis in specific instances. Before going further, we need to consider some semantic points. When it was first suggested that the secretory response of pancreatic islet beta cells to D-glucose is causally linked to an accelerated catabolism of the hexose, the term ‘glucoreceptor’ was used to refer to an essential component (e.g. glucokinase) of the biochemical machinery that regulates D-glucose metabolism in these cells [4, 5]. Incidentally, the view that glucokinase would act as a glucoreceptor in pancreatic beta cells was refuted, considering the relevance of regulatory steps distal to glucose phosphorylation in the control of glucose catabolism in islet cells [6]. Further, the word glucoreceptor is also often used when referring to the glucose-sensing device in pancreatic islet cells, regardless of the intimate mode of action of the hexose in islet cells [7]. A possible reconciliation between the fuel and receptor hypotheses has been suggested, based on the following considerations. First, a component of the beta cell boundary that is stereospecific for D-glucose, as distinct from L-glucose, exists in the form of the carrier mediating glucose transport across the plasma membrane [8, 9]. Second, the possibility remains that the D-glucose molecule itself, independently from its role as substrate, modulates the activity of enzymes involved in the metabolism of carbohydrates in the islet cells [7]. For instance, it was documented that the binding of D-glucose to the active phosphorylated a-form of glycogen phosphorylase decreases the activity of this enzyme and renders it a better substrate for inactivation by phosphorylase phosphatase [10]. Such a
Diabetologia
process is implicated in two major features of the phenomenon of glucotoxicity in insulin-producing cells, i.e. the paradoxical early and transient inhibition of insulin output in response to a rise in extracellular D-glucose concentration, and the altered anomeric specificity of glucose-stimulated insulin secretion [11–14]. Last, the non-metabolised analogue of L-leucine, 2-amino-bicyclo[2, 2, 1]heptane-2-carboxylic acid, was found to stimulate insulin release by causing allosteric activation of glutamate dehydrogenase which could, therefore, be looked upon as a stereospecific receptor for branched chain amino acids [15]. With this background information in mind, attention will now be drawn to true receptors involved in the process of insulin secretion. The possible presence of a sweet taste receptor in insulinproducing islet cells was first hypothesised in light of the parallel perturbation of the anomeric specificity in the process of glucose-induced insulin secretion [16] and in the recognition of the sweet taste of these anomers [17, 18] in individuals with diabetes mellitus. Incidentally, the altered anomeric specificity of glucose-induced insulin release was eventually ascribed to the accumulation of glycogen in pancreatic beta cells, it being referred to as an anomeric malaise [11]. Nevertheless, in the light of the stimulation of insulin release by artificial sweeteners, attention was again drawn to the possible involvement of either sweet taste receptors or, considering the bitter taste or aftertaste of some artificial sweeteners, bitter taste receptors in their insulinotropic action [19]. The latter proposal was further considered in the light of both the stimulation of insulin release by the α- and β-anomers of glucose pentaacetate and their bitter taste [20, 21]. As previously reviewed in detail [20], the effects of β-Lglucose pentaacetate on such variables as electrical activity of pancreatic islet beta cells [22], 86Rb outflow from prelablelled islets [23], cytosolic Ca2+ concentration in mouse islets [22], and the release of insulin, somatostatin and glucagon from the perfused rat pancreas [24] are all compatible with the suggestion that this ester directly interacts with a receptor leading to plasma membrane depolarisation, induction of electrical activity and increase in cytosolic Ca2+ concentration. In the light of these findings, the possible presence in purified beta cells of the α-gustducin G-protein, which is known to be involved in the process by which taste buds identify bitter compounds, was also investigated [20]. In this regard, however, our preliminary findings remain to be confirmed by other groups, although it does appear that bitter compounds stimulate insulin secretion: as a matter of fact, denatonium, one of the most bitter-tasting substances known, was found to stimulate insulin secretion in both clonal HIT-T15 beta cells and rat pancreatic islets [25]. This insulinotropic action, documented in the presence of 8.3 mmol/l D-glucose, was abolished in the absence of extracellular Ca2+ or in the presence of the Ca2+-channel blocker nitrendipine and inhibited by the α2-adrenergic agonist
clonidine. Furthermore, it could not be attributed to any direct effect on voltage-gated calcium channels or cellular cyclic AMP levels, and no evidence was found to suggest activation by denatonium of either gustducin or transducin in the beta cells. The insulinotropic action of this bitter compound was eventually ascribed to its interaction with ATP-responsive K+ channels, leading to the depolarisation of beta cells and resulting increase in Ca2+ influx [25]. In an impressive series of recent publications, I. Kojima and colleagues revealed the presence of a sweet taste receptor in pancreatic islet beta cells and documented its multimodal signalling process in these cells [26–29]. Three main messages emerge from these investigations. First, the expression of the sweet taste receptor T1R3 in both MIN6 cells and mouse pancreatic islets was documented by RT-PCR and immunohistochemistry. The activation of this receptor by the artificial sweetener sucralose also stimulated insulin secretion in both MIN6 cells and mouse pancreatic islets. The artificial sweeteners saccharin and acesulfame K also increased insulin release from MIN6 cells at both low (3 mmol/l) and high (25 mmol/l) glucose concentration. The secretory response to sucralose was concentration-related in the 3 to 30 mmol/l range. Sucralose also induced elevation of both cytoplasmic cyclic AMP and Ca2+ concentrations. The sucralose-induced elevation of cytosolic Ca2+ was inhibited by gumarin, an antagonist of the sweet taste receptor, when extracellular Ca2+ was removed, or in the presence of nifedipine, an inhibitor of the L-type voltage-dependent calcium channel. As a rule, parallel changes were observed for cytoplasmic cyclic AMP concentration, suggesting a possible role for the calciumdependent activation of adenylate cyclase, e.g. at the intervention of calmodulin. It was proposed that the sucralose-induced increase in cytosolic Ca2+ may also involve release of Ca2+ from an intracellular pool, as an inhibitor of inositol (1,4,5)-trisphosphate receptor nearly completely blocked sucralose-induced elevation of cytosolic Ca2+ concentration. Finally, sucralose was found to activate protein kinase C [26, 27]. Second, in a further study, the effects of distinct artificial sweeteners (sucralose, saccharin, acesulfame K, glycyrrhizin and N-methyl-D-glucamine) on both cytoplasmic Ca2+ and cyclic AMP concentrations were assessed in MIN6 cells. The results revealed that these various types of sweetener activate the sweet taste receptor differently, generating distinct patterns of intracellular signals. Hence, it was concluded that the sweet taste receptor has amazing multimodal functions [28]. Last, and most importantly, the same authors provided evidence that glucose promotes its own metabolism by acting on the cell-surface glucose-sensing receptor T1R3. Based mainly on the measurement of intracellular ATP in MIN6 cells transfected with the luciferase gene, their research documented the effect of glucose to increase intracellular ATP (which coincided with an increase in the intracellular ATP/ADP ratio) and its inhibition by 2-4-dinitrophenol, indicating a role for
Diabetologia
mitochondria in maintaining ATP levels, or by 2-cyclohexen1-one, an inhibitor of glucokinase. The artificial sweetener sucralose, which is not metabolised in beta cells, also increased intracellular ATP in a dose-dependent manner (0.5 to 10.0 mmol/l) and interacted with glucose in more than a mere additive manner. Sucralose was proposed to promote mitochondrial metabolism, acting in a synergistic manner with methyl-succinate, a cell-permeable analogue of succinate used as mitochondrial fuel. The view that glucose may activate the T1R3 receptor was supported by the finding that a nonmetabolised analogue of glucose, namely 3-O-methyl-glucose, also caused a rapid and transient increase in intracellular ATP. Even 2-deoxy-D-glucose, a non-metabolised analogue of glucose that also inhibits glucokinase activity, caused in the presence of 5.5 mmol/l glucose first a rapid and transient increase in ATP, followed by a graded decrease, the intracellular ATP eventually reaching values below the basal level. Moreover 3-O-methyl-glucose acted synergistically with methylsuccinate, indicating that the glucose analogue augmented succinate metabolism. Last, in T1R3 knocked-down cells, obtained by adding siRNA (which reduced the expression of T1R3 by about 70%), the effect of either glucose or sucralose on intracellular ATP was impaired. Glucose-induced insulin secretion was also significantly attenuated in the T1R3 knocked-down cells [29]. The possible extension of this knowledge to other types of pancreatic endocrine cells, e.g. glucagon-producing cells, remains an open question presently under investigation. In summary, activation of the T1R3 cell-surface glucosesensing receptor by sucralose, like the stimulation of beta cells by D-glucose, leads to increases in cytoplasmic Ca2+ and cyclic AMP and promotion of glucose metabolism (see the text box). These findings raise the possibility that the perturbation in the recognition of the sweet taste of D-glucose anomers in diabetic patients, as already mentioned in this review, may coincide with an impaired response of the sweet taste receptor present in beta cells and, hence, may participate in the impairment of glucose-stimulated insulin secretion. Common effects of sucralose and D-glucose in beta cells • Promotion of mitochondrial metabolism • Increase in intracellular ATP concentration • Gating of voltage-dependent calcium channels • Release of Ca2+ from an intracellular pool caused by stimulation of phosphoinositide hydrolysis • Increase of cytosolic Ca2+ concentration • Activation of adenylate cyclase with increase in cytoplasmic cyclic AMP concentration • Activation of protein kinase C • Stimulation of insulin release
A further instance in which both the fuel hypothesis and the receptor hypothesis for the stimulation of insulin release may apply refers to the insulinotropic action of D-fructose. As recently and extensively reviewed, several of the metabolic, functional and pathological aspects of the process of D -fructose-induced insulin secretion suggest that the insulinotropic action of this ketohexose cannot be fully accounted for by its catabolic fate in pancreatic islet cells [30–32]. First, despite the fact that the cationic and insulinotropic effects of D-glucose and/or D-fructose appear tightly related to the cytosolic ATP/ADP ratio [33] and may involve tight and presumably reciprocal coupling between mitochondrial oxidative processes and ATP-consuming functional events [34], an obvious dissociation between oxidative and secretory variables was observed when comparing the situations found in islets exposed to 8.3 mmol/l D-glucose, on the one hand, and the combination of 6.0 mmol/l D-glucose and 80.0 mmol/l D-fructose, on the other hand [35]. Second, a lack of parallels between the biosynthetic and secretory responses of pancreatic islets to D-fructose used in high concentrations (80.0 to 240.0 mmol/l) again supported the view that the insulinotropic action of the ketohexose does not entail the same metabolic determinants as those operative in D-glucosestimulated islets [36]. Last, from a pathological perspective, an impairment of the secretory response to D-fructose that could apparently not be attributed to any obvious defect in D-fructose catabolism was identified in islets from Goto-Kakizaki rats, a current animal model of inherited type 2 diabetes mellitus [37]. Likewise, in pancreatic islets from adult rats injected with streptozotocin during the neonatal period (STZ rats), a preferential suppression of the secretory response to D-fructose, compared with D-glucose, was observed despite close analogies between control and STZ rats in terms of the respective catabolic fates of D-glucose and D-fructose, as well as their reciprocal metabolic effects [38]. Thus, the latter observation again supports the idea that the insulinotropic action of D-fructose may not be entirely accounted for by its nutritional value in islet cells. Consistent with these findings, it was recently proposed that sweet taste receptor signalling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion [39]. Thus, at variance with the situation found in control mouse islets, D-fructose (10.0 mmol/l) failed, in the presence of D-glucose (8.3 mmol/l), to increase intracellular Ca2+ concentration and to induce insulin release from T1r2 knockout mouse islets, whilst no significant difference in glucose-induced Ca 2+ response or insulin release was observed between islets from wild-type and T1r2 knockout mice. Likewise, in vivo, the intravenous administration of D-fructose, which induced an early transient increase of plasma insulin concentration in wild-type mice, failed to do so in T1r2 knockout mice. Evidence was obtained that D-fructose activates phospholipase C in pancreatic islets, even in the
Diabetologia
presence of nifedipine or absence of extracellular Ca2+. The resulting activation of the calcium-sensitive cation channel TRPM5 was proposed as a convergence point between sweet taste receptor- and glucose-induced insulin secretion in beta cells. In conclusion, this review illustrates that the fuel hypothesis and the receptor hypothesis concerning the insulinotropic action of selected nutrient secretagogues are not necessarily incompatible with one another. Duality of interest The author declares that there is no duality of interest associated with this manuscript. Contribution statement WJM was responsible for the conception and design of the manuscript.
References 1. Malaisse WJ, Sener A, Herchuelz A, Hutton JC (1979) Insulin release: the fuel hypothesis. Metabolism 28:373–386 2. Malaisse WJ (1983) Insulin release: the fuel concept. Diabete Metab 9:313–320 3. Matschinsky FM, Ellerman J, Stillings S, Raybaud F, Pace C, Zawalich W (1975) Hexoses and insulin secretion. In: Hasselblatt A, Bruchhausen FV (eds) Insulin (part 2). Springer, Heidelberg, pp 79–114 4. Meglasson MD, Matschinsky FM (1984) New perspectives on pancreatic islet glucokinase. Am J Physiol 246:E1–E13 5. Garfinkel D, Garfinkel L, Meglasson MD, Matschinsky FM (1984) Computer modelling identifies glucokinase as glucose sensor of pancreatic B cell. Am J Physiol 247:R527–R536 6. Malaisse WJ, Sener A (1985) Glucokinase is not the pancreatic B cell glucoreceptor. Diabetologia 28:520–527 7. Malaisse WJ (1987) Insulin release: the glucoreceptor myth. Med Sci Res 15:65–67 8. Hellman B, Sehlin J, Täljedal IB (1971) Evidence for mediated transport of glucose in mammalian pancreatic B cells. Biochim Biophys Acta 241:147–154 9. Price S (1973) Pancreatic islet membranes: extraction of a possible glucoreceptor. Biochim Biophys Acta 318:459–463 10. Bollen M, Malaisse-Lagae F, Malaisse W, Stalmans W (1990) The interaction of phosphorylase a with D-glucose displays α-stereospecificity. Biochim Biophys Acta 1038:141–145 11. Malaisse WJ (1991) The anomeric malaise: a manifestation of B cell glucotoxicity. Horm Metab Res 23:307–311 12. Marynissen G, Leclercq-Meyer V, Sener A, Malaisse WJ (1990) Perturbation of pancreatic islet function in glucose-infused rats. Metabolism 39:87–95 13. Malaisse WJ, Marynissen G, Sener A (1992) Possible role of glycogen accumulation in B cell glucotoxicity. Metabolism 41: 814–819 14. Malaisse WJ, Maggetto C, Leclercq-Meyer V, Sener A (1993) Interference of glycogenolysis with glycolysis in pancreatic islets from glucose-infused rats. J Clin Invest 91:432–436 15. Malaisse WJ, Sener A, Malaisse-Lagae F (1981) Insulin release: reconciliation of the receptor and metabolic hypotheses. Nutrient receptors in islet cells. Mol Cell Biochem 37:157–165 16. Rovira A, Garrotte FJ, Valverde I, Malaisse WJ (1987) Anomeric specificity of glucose-induced insulin release in normal and diabetic subjects. Diabetes Res 5:119–124 17. Niki A, Niki H (1975) Is diabetes mellitus a disorder of the glucoreceptor? Lancet 2:658
18. Malaisse-Lagae F, Malaisse WJ (1986) Abnormal identification of the sweet taste of D-glucose anomers. Diabetologia 29:344–345 (Letter) 19. Malaisse WJ, Vanonderbergen A, Louchami K, Jijakli H, Malaisse-Lagae F (1998) Effects of artificial sweeteners on insulin release and cationic fluxes in rat pancreatic islets. Cell Signal 10:727–733 20. Malaisse WJ (1998) The riddle of L -glucose pentaacetate insulinotropic action. Int J Mol Med 2:383–388 21. Malaisse WJ, Malaisse-Lagae F (1997) Bitter taste of monosaccharide pentaacetate esters. Biochem Mol Biol Int 43:1367–1371 22. Pomares R, Ropero AB, Sanchez-Andres JV, Nadal A, Soria B, Malaisse WJ (1999) Effect of hexose pentaacetates on electrical activity and cytosolic Ca2+ in mouse pancreatic islets. Int J Mol Med 3:15–20 23. Malaisse WJ, Best LC, Herchuelz A, et al (1998) Insulinotropic action of β-L-glucose pentaacetate. Am J Physiol 275:E993-El006. 24. Leclercq-Meyer V, Malaisse WJ (1998) Dual mode of action of glucose pentaacetates on hormonal secretion from the isolated perfused rat pancreas. Am J Physiol 275:E610–E617 25. Straub SG, Mulvaney-Musa J, Yajima H, Weiland GA, Sharp GWG (2003) Stimulation of insulin secretion by denatonium, one of the most bitter-tasting substances known. Diabetes 52:356–364 26. Nakagawa Y, Nagasawa M, Yamada S et al (2009) Sweet taste receptor expressed in pancreatic beta-cells activate the calcium and cyclic AMP signalling systems and stimulates insulin secretion. PLoS ONE 4:e5106 27. Kojima I, Nakagawa Y (2011) The role of the sweet taste receptor in enteroendocrine cells and pancreatic β-cells. Diabetes Metab J 34: 451–457 28. Nakagawa Y, Nagasawa M, Mogami H, Lohse M, Ninomiya Y, Kojima I (2013) Multinodal function of the sweet taste receptor expressed in pancreatic β-cells: generation of diverse patterns of intracellular signals by sweet agonists. Endocr J 60:1191–1206 29. Nakagawa Y, Ohtsu Y, Nagasawa M, Shibata H, Kojima I (2013) Glucose promotes its own metabolism by acting on the cell-surface glucose-sensing receptor T1R3. Endocr J 10:1507 30. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: metabolic aspects. Curr Top Biochem Res (in press). 31. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: functional aspects. Curr Top Biochem Res (in press). 32. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: pathological aspects. Curr Top Biochem Res (in press). 33. Giroix M-H, Agascioglu E, Oguzhan B et al (2006) Opposite effects of D-fructose on total versus cytosolic ATP/ADP ratio in rat pancreatic islet cells. Biochim Biophys Acta 1757:773–780 34. Sener A, Blachier F, Malaisse WJ (1990) Hexose metabolism in pancreatic islets: comparison and interaction between D-glucose and D-fructose. Turk J Med Biol Res 1:5–12 35. Sener A, Malaisse WJ (1996) Hexose metabolism in pancreatic islets: apparent dissociation between the secretory and metabolic effects of D-fructose. Biochem Mol Med 59:182–186 36. Viñambres C, Villanueva-Peñacarrillo ML, Valverde I, Malaisse WJ (1997) Failure of D-fructose to stimulate protein biosynthesis in pancreatic islets. Biochem Mol Biol Int 41:571–574 37. Giroix M-H, Scruel O, Ladrière L, Sener A, Portha B, Malaisse WJ (1999) Metabolic and secretory interactions between D-glucose and D-fructose in islets from GK rats. Endocrinology 140:5556–5565 38. Scruel O, Giroix M-H, Sener A, Portha B, Malaisse WJ (1999) Metabolic and secretory response to D-fructose in islets from adult rats injected with streptozotocin during the neonatal period. Mol Genet Metab 68:86–90 39. Kyriazis GA, Soundarapandian MM, Tyrberg B (2012) Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion. Proc Natl Acad Sci U S A 109:E524–E532
REVIEW
Insulin release: the receptor hypothesis Willy J. Malaisse
Received: 16 January 2014 / Accepted: 28 February 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract It is currently believed that the stimulation of insulin release by nutrient secretagogues reflects their capacity to act as fuel in pancreatic islet beta cells. In this review, it is proposed that such a fuel concept is not incompatible with a receptor hypothesis postulating the participation of cellsurface receptors in the recognition of selected nutrients as insulinotropic agents. Pursuant to this, attention is drawn to such matters as the anomeric specificity of the beta cell secretory response to D-glucose and its perturbation in diabetes mellitus, the insulinotropic action of artificial sweeteners, the possible role of bitter taste receptors in the stimulation of insulin secretion by L-glucose pentaacetate, the recently documented presence of cell-surface sweet taste receptors in insulinproducing cells, the multimodal signalling process resulting from the activation of these latter receptors, and the presence in beta cells of a sweet taste receptor mediating the fructoseinduced potentiation of glucose-stimulated insulin secretion. Keywords Bitter taste receptor . Insulin release . Pancreatic islet beta cells . Sweet taste receptor In an article entitled ‘Insulin release: the fuel hypothesis’ published in 1979, it was proposed that the capacity of nutrients to act as fuel in pancreatic islet cells may account for their effects on the secretion of insulin and possibly other pancreatic hormones [1]. This fuel concept was mainly based on such findings as the parallels between the metabolic fate and insulinotropic action of several carbohydrates, including the α- and β-anomers of D-glucose or D-mannose, and the stimulation of insulin evoked either in the absence of extracellular D-glucose by glycogenolysis and glycolysis from endogenous stores of W. J. Malaisse (*) Department of Biochemistry, Université Libre de Bruxelles, 808 Route de Lennik, 1070 Brussels, Belgium e-mail: [email protected]
glycogen, by the poorly oxidised analogue of pyruvate, 3-phenylpyruvate, or by selected keto and amino acids, including the analogue of L-leucine, 2-amino-bicyclo[2,2,1]heptane-2carboxylic acid [1, 2]. The fuel concept obviously differs from the view that nutrient secretagogues are recognised by islet cells through the intervention of stereospecific receptors, possibly located at the cell membrane [3]. The main aim of this review is to illustrate the pertinence of the receptor hypothesis in specific instances. Before going further, we need to consider some semantic points. When it was first suggested that the secretory response of pancreatic islet beta cells to D-glucose is causally linked to an accelerated catabolism of the hexose, the term ‘glucoreceptor’ was used to refer to an essential component (e.g. glucokinase) of the biochemical machinery that regulates D-glucose metabolism in these cells [4, 5]. Incidentally, the view that glucokinase would act as a glucoreceptor in pancreatic beta cells was refuted, considering the relevance of regulatory steps distal to glucose phosphorylation in the control of glucose catabolism in islet cells [6]. Further, the word glucoreceptor is also often used when referring to the glucose-sensing device in pancreatic islet cells, regardless of the intimate mode of action of the hexose in islet cells [7]. A possible reconciliation between the fuel and receptor hypotheses has been suggested, based on the following considerations. First, a component of the beta cell boundary that is stereospecific for D-glucose, as distinct from L-glucose, exists in the form of the carrier mediating glucose transport across the plasma membrane [8, 9]. Second, the possibility remains that the D-glucose molecule itself, independently from its role as substrate, modulates the activity of enzymes involved in the metabolism of carbohydrates in the islet cells [7]. For instance, it was documented that the binding of D-glucose to the active phosphorylated a-form of glycogen phosphorylase decreases the activity of this enzyme and renders it a better substrate for inactivation by phosphorylase phosphatase [10]. Such a
Diabetologia
process is implicated in two major features of the phenomenon of glucotoxicity in insulin-producing cells, i.e. the paradoxical early and transient inhibition of insulin output in response to a rise in extracellular D-glucose concentration, and the altered anomeric specificity of glucose-stimulated insulin secretion [11–14]. Last, the non-metabolised analogue of L-leucine, 2-amino-bicyclo[2, 2, 1]heptane-2-carboxylic acid, was found to stimulate insulin release by causing allosteric activation of glutamate dehydrogenase which could, therefore, be looked upon as a stereospecific receptor for branched chain amino acids [15]. With this background information in mind, attention will now be drawn to true receptors involved in the process of insulin secretion. The possible presence of a sweet taste receptor in insulinproducing islet cells was first hypothesised in light of the parallel perturbation of the anomeric specificity in the process of glucose-induced insulin secretion [16] and in the recognition of the sweet taste of these anomers [17, 18] in individuals with diabetes mellitus. Incidentally, the altered anomeric specificity of glucose-induced insulin release was eventually ascribed to the accumulation of glycogen in pancreatic beta cells, it being referred to as an anomeric malaise [11]. Nevertheless, in the light of the stimulation of insulin release by artificial sweeteners, attention was again drawn to the possible involvement of either sweet taste receptors or, considering the bitter taste or aftertaste of some artificial sweeteners, bitter taste receptors in their insulinotropic action [19]. The latter proposal was further considered in the light of both the stimulation of insulin release by the α- and β-anomers of glucose pentaacetate and their bitter taste [20, 21]. As previously reviewed in detail [20], the effects of β-Lglucose pentaacetate on such variables as electrical activity of pancreatic islet beta cells [22], 86Rb outflow from prelablelled islets [23], cytosolic Ca2+ concentration in mouse islets [22], and the release of insulin, somatostatin and glucagon from the perfused rat pancreas [24] are all compatible with the suggestion that this ester directly interacts with a receptor leading to plasma membrane depolarisation, induction of electrical activity and increase in cytosolic Ca2+ concentration. In the light of these findings, the possible presence in purified beta cells of the α-gustducin G-protein, which is known to be involved in the process by which taste buds identify bitter compounds, was also investigated [20]. In this regard, however, our preliminary findings remain to be confirmed by other groups, although it does appear that bitter compounds stimulate insulin secretion: as a matter of fact, denatonium, one of the most bitter-tasting substances known, was found to stimulate insulin secretion in both clonal HIT-T15 beta cells and rat pancreatic islets [25]. This insulinotropic action, documented in the presence of 8.3 mmol/l D-glucose, was abolished in the absence of extracellular Ca2+ or in the presence of the Ca2+-channel blocker nitrendipine and inhibited by the α2-adrenergic agonist
clonidine. Furthermore, it could not be attributed to any direct effect on voltage-gated calcium channels or cellular cyclic AMP levels, and no evidence was found to suggest activation by denatonium of either gustducin or transducin in the beta cells. The insulinotropic action of this bitter compound was eventually ascribed to its interaction with ATP-responsive K+ channels, leading to the depolarisation of beta cells and resulting increase in Ca2+ influx [25]. In an impressive series of recent publications, I. Kojima and colleagues revealed the presence of a sweet taste receptor in pancreatic islet beta cells and documented its multimodal signalling process in these cells [26–29]. Three main messages emerge from these investigations. First, the expression of the sweet taste receptor T1R3 in both MIN6 cells and mouse pancreatic islets was documented by RT-PCR and immunohistochemistry. The activation of this receptor by the artificial sweetener sucralose also stimulated insulin secretion in both MIN6 cells and mouse pancreatic islets. The artificial sweeteners saccharin and acesulfame K also increased insulin release from MIN6 cells at both low (3 mmol/l) and high (25 mmol/l) glucose concentration. The secretory response to sucralose was concentration-related in the 3 to 30 mmol/l range. Sucralose also induced elevation of both cytoplasmic cyclic AMP and Ca2+ concentrations. The sucralose-induced elevation of cytosolic Ca2+ was inhibited by gumarin, an antagonist of the sweet taste receptor, when extracellular Ca2+ was removed, or in the presence of nifedipine, an inhibitor of the L-type voltage-dependent calcium channel. As a rule, parallel changes were observed for cytoplasmic cyclic AMP concentration, suggesting a possible role for the calciumdependent activation of adenylate cyclase, e.g. at the intervention of calmodulin. It was proposed that the sucralose-induced increase in cytosolic Ca2+ may also involve release of Ca2+ from an intracellular pool, as an inhibitor of inositol (1,4,5)-trisphosphate receptor nearly completely blocked sucralose-induced elevation of cytosolic Ca2+ concentration. Finally, sucralose was found to activate protein kinase C [26, 27]. Second, in a further study, the effects of distinct artificial sweeteners (sucralose, saccharin, acesulfame K, glycyrrhizin and N-methyl-D-glucamine) on both cytoplasmic Ca2+ and cyclic AMP concentrations were assessed in MIN6 cells. The results revealed that these various types of sweetener activate the sweet taste receptor differently, generating distinct patterns of intracellular signals. Hence, it was concluded that the sweet taste receptor has amazing multimodal functions [28]. Last, and most importantly, the same authors provided evidence that glucose promotes its own metabolism by acting on the cell-surface glucose-sensing receptor T1R3. Based mainly on the measurement of intracellular ATP in MIN6 cells transfected with the luciferase gene, their research documented the effect of glucose to increase intracellular ATP (which coincided with an increase in the intracellular ATP/ADP ratio) and its inhibition by 2-4-dinitrophenol, indicating a role for
Diabetologia
mitochondria in maintaining ATP levels, or by 2-cyclohexen1-one, an inhibitor of glucokinase. The artificial sweetener sucralose, which is not metabolised in beta cells, also increased intracellular ATP in a dose-dependent manner (0.5 to 10.0 mmol/l) and interacted with glucose in more than a mere additive manner. Sucralose was proposed to promote mitochondrial metabolism, acting in a synergistic manner with methyl-succinate, a cell-permeable analogue of succinate used as mitochondrial fuel. The view that glucose may activate the T1R3 receptor was supported by the finding that a nonmetabolised analogue of glucose, namely 3-O-methyl-glucose, also caused a rapid and transient increase in intracellular ATP. Even 2-deoxy-D-glucose, a non-metabolised analogue of glucose that also inhibits glucokinase activity, caused in the presence of 5.5 mmol/l glucose first a rapid and transient increase in ATP, followed by a graded decrease, the intracellular ATP eventually reaching values below the basal level. Moreover 3-O-methyl-glucose acted synergistically with methylsuccinate, indicating that the glucose analogue augmented succinate metabolism. Last, in T1R3 knocked-down cells, obtained by adding siRNA (which reduced the expression of T1R3 by about 70%), the effect of either glucose or sucralose on intracellular ATP was impaired. Glucose-induced insulin secretion was also significantly attenuated in the T1R3 knocked-down cells [29]. The possible extension of this knowledge to other types of pancreatic endocrine cells, e.g. glucagon-producing cells, remains an open question presently under investigation. In summary, activation of the T1R3 cell-surface glucosesensing receptor by sucralose, like the stimulation of beta cells by D-glucose, leads to increases in cytoplasmic Ca2+ and cyclic AMP and promotion of glucose metabolism (see the text box). These findings raise the possibility that the perturbation in the recognition of the sweet taste of D-glucose anomers in diabetic patients, as already mentioned in this review, may coincide with an impaired response of the sweet taste receptor present in beta cells and, hence, may participate in the impairment of glucose-stimulated insulin secretion. Common effects of sucralose and D-glucose in beta cells • Promotion of mitochondrial metabolism • Increase in intracellular ATP concentration • Gating of voltage-dependent calcium channels • Release of Ca2+ from an intracellular pool caused by stimulation of phosphoinositide hydrolysis • Increase of cytosolic Ca2+ concentration • Activation of adenylate cyclase with increase in cytoplasmic cyclic AMP concentration • Activation of protein kinase C • Stimulation of insulin release
A further instance in which both the fuel hypothesis and the receptor hypothesis for the stimulation of insulin release may apply refers to the insulinotropic action of D-fructose. As recently and extensively reviewed, several of the metabolic, functional and pathological aspects of the process of D -fructose-induced insulin secretion suggest that the insulinotropic action of this ketohexose cannot be fully accounted for by its catabolic fate in pancreatic islet cells [30–32]. First, despite the fact that the cationic and insulinotropic effects of D-glucose and/or D-fructose appear tightly related to the cytosolic ATP/ADP ratio [33] and may involve tight and presumably reciprocal coupling between mitochondrial oxidative processes and ATP-consuming functional events [34], an obvious dissociation between oxidative and secretory variables was observed when comparing the situations found in islets exposed to 8.3 mmol/l D-glucose, on the one hand, and the combination of 6.0 mmol/l D-glucose and 80.0 mmol/l D-fructose, on the other hand [35]. Second, a lack of parallels between the biosynthetic and secretory responses of pancreatic islets to D-fructose used in high concentrations (80.0 to 240.0 mmol/l) again supported the view that the insulinotropic action of the ketohexose does not entail the same metabolic determinants as those operative in D-glucosestimulated islets [36]. Last, from a pathological perspective, an impairment of the secretory response to D-fructose that could apparently not be attributed to any obvious defect in D-fructose catabolism was identified in islets from Goto-Kakizaki rats, a current animal model of inherited type 2 diabetes mellitus [37]. Likewise, in pancreatic islets from adult rats injected with streptozotocin during the neonatal period (STZ rats), a preferential suppression of the secretory response to D-fructose, compared with D-glucose, was observed despite close analogies between control and STZ rats in terms of the respective catabolic fates of D-glucose and D-fructose, as well as their reciprocal metabolic effects [38]. Thus, the latter observation again supports the idea that the insulinotropic action of D-fructose may not be entirely accounted for by its nutritional value in islet cells. Consistent with these findings, it was recently proposed that sweet taste receptor signalling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion [39]. Thus, at variance with the situation found in control mouse islets, D-fructose (10.0 mmol/l) failed, in the presence of D-glucose (8.3 mmol/l), to increase intracellular Ca2+ concentration and to induce insulin release from T1r2 knockout mouse islets, whilst no significant difference in glucose-induced Ca 2+ response or insulin release was observed between islets from wild-type and T1r2 knockout mice. Likewise, in vivo, the intravenous administration of D-fructose, which induced an early transient increase of plasma insulin concentration in wild-type mice, failed to do so in T1r2 knockout mice. Evidence was obtained that D-fructose activates phospholipase C in pancreatic islets, even in the
Diabetologia
presence of nifedipine or absence of extracellular Ca2+. The resulting activation of the calcium-sensitive cation channel TRPM5 was proposed as a convergence point between sweet taste receptor- and glucose-induced insulin secretion in beta cells. In conclusion, this review illustrates that the fuel hypothesis and the receptor hypothesis concerning the insulinotropic action of selected nutrient secretagogues are not necessarily incompatible with one another. Duality of interest The author declares that there is no duality of interest associated with this manuscript. Contribution statement WJM was responsible for the conception and design of the manuscript.
References 1. Malaisse WJ, Sener A, Herchuelz A, Hutton JC (1979) Insulin release: the fuel hypothesis. Metabolism 28:373–386 2. Malaisse WJ (1983) Insulin release: the fuel concept. Diabete Metab 9:313–320 3. Matschinsky FM, Ellerman J, Stillings S, Raybaud F, Pace C, Zawalich W (1975) Hexoses and insulin secretion. In: Hasselblatt A, Bruchhausen FV (eds) Insulin (part 2). Springer, Heidelberg, pp 79–114 4. Meglasson MD, Matschinsky FM (1984) New perspectives on pancreatic islet glucokinase. Am J Physiol 246:E1–E13 5. Garfinkel D, Garfinkel L, Meglasson MD, Matschinsky FM (1984) Computer modelling identifies glucokinase as glucose sensor of pancreatic B cell. Am J Physiol 247:R527–R536 6. Malaisse WJ, Sener A (1985) Glucokinase is not the pancreatic B cell glucoreceptor. Diabetologia 28:520–527 7. Malaisse WJ (1987) Insulin release: the glucoreceptor myth. Med Sci Res 15:65–67 8. Hellman B, Sehlin J, Täljedal IB (1971) Evidence for mediated transport of glucose in mammalian pancreatic B cells. Biochim Biophys Acta 241:147–154 9. Price S (1973) Pancreatic islet membranes: extraction of a possible glucoreceptor. Biochim Biophys Acta 318:459–463 10. Bollen M, Malaisse-Lagae F, Malaisse W, Stalmans W (1990) The interaction of phosphorylase a with D-glucose displays α-stereospecificity. Biochim Biophys Acta 1038:141–145 11. Malaisse WJ (1991) The anomeric malaise: a manifestation of B cell glucotoxicity. Horm Metab Res 23:307–311 12. Marynissen G, Leclercq-Meyer V, Sener A, Malaisse WJ (1990) Perturbation of pancreatic islet function in glucose-infused rats. Metabolism 39:87–95 13. Malaisse WJ, Marynissen G, Sener A (1992) Possible role of glycogen accumulation in B cell glucotoxicity. Metabolism 41: 814–819 14. Malaisse WJ, Maggetto C, Leclercq-Meyer V, Sener A (1993) Interference of glycogenolysis with glycolysis in pancreatic islets from glucose-infused rats. J Clin Invest 91:432–436 15. Malaisse WJ, Sener A, Malaisse-Lagae F (1981) Insulin release: reconciliation of the receptor and metabolic hypotheses. Nutrient receptors in islet cells. Mol Cell Biochem 37:157–165 16. Rovira A, Garrotte FJ, Valverde I, Malaisse WJ (1987) Anomeric specificity of glucose-induced insulin release in normal and diabetic subjects. Diabetes Res 5:119–124 17. Niki A, Niki H (1975) Is diabetes mellitus a disorder of the glucoreceptor? Lancet 2:658
18. Malaisse-Lagae F, Malaisse WJ (1986) Abnormal identification of the sweet taste of D-glucose anomers. Diabetologia 29:344–345 (Letter) 19. Malaisse WJ, Vanonderbergen A, Louchami K, Jijakli H, Malaisse-Lagae F (1998) Effects of artificial sweeteners on insulin release and cationic fluxes in rat pancreatic islets. Cell Signal 10:727–733 20. Malaisse WJ (1998) The riddle of L -glucose pentaacetate insulinotropic action. Int J Mol Med 2:383–388 21. Malaisse WJ, Malaisse-Lagae F (1997) Bitter taste of monosaccharide pentaacetate esters. Biochem Mol Biol Int 43:1367–1371 22. Pomares R, Ropero AB, Sanchez-Andres JV, Nadal A, Soria B, Malaisse WJ (1999) Effect of hexose pentaacetates on electrical activity and cytosolic Ca2+ in mouse pancreatic islets. Int J Mol Med 3:15–20 23. Malaisse WJ, Best LC, Herchuelz A, et al (1998) Insulinotropic action of β-L-glucose pentaacetate. Am J Physiol 275:E993-El006. 24. Leclercq-Meyer V, Malaisse WJ (1998) Dual mode of action of glucose pentaacetates on hormonal secretion from the isolated perfused rat pancreas. Am J Physiol 275:E610–E617 25. Straub SG, Mulvaney-Musa J, Yajima H, Weiland GA, Sharp GWG (2003) Stimulation of insulin secretion by denatonium, one of the most bitter-tasting substances known. Diabetes 52:356–364 26. Nakagawa Y, Nagasawa M, Yamada S et al (2009) Sweet taste receptor expressed in pancreatic beta-cells activate the calcium and cyclic AMP signalling systems and stimulates insulin secretion. PLoS ONE 4:e5106 27. Kojima I, Nakagawa Y (2011) The role of the sweet taste receptor in enteroendocrine cells and pancreatic β-cells. Diabetes Metab J 34: 451–457 28. Nakagawa Y, Nagasawa M, Mogami H, Lohse M, Ninomiya Y, Kojima I (2013) Multinodal function of the sweet taste receptor expressed in pancreatic β-cells: generation of diverse patterns of intracellular signals by sweet agonists. Endocr J 60:1191–1206 29. Nakagawa Y, Ohtsu Y, Nagasawa M, Shibata H, Kojima I (2013) Glucose promotes its own metabolism by acting on the cell-surface glucose-sensing receptor T1R3. Endocr J 10:1507 30. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: metabolic aspects. Curr Top Biochem Res (in press). 31. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: functional aspects. Curr Top Biochem Res (in press). 32. Malaisse WJ (2014) D-fructose metabolism and insulinotropic action in pancreatic islets: pathological aspects. Curr Top Biochem Res (in press). 33. Giroix M-H, Agascioglu E, Oguzhan B et al (2006) Opposite effects of D-fructose on total versus cytosolic ATP/ADP ratio in rat pancreatic islet cells. Biochim Biophys Acta 1757:773–780 34. Sener A, Blachier F, Malaisse WJ (1990) Hexose metabolism in pancreatic islets: comparison and interaction between D-glucose and D-fructose. Turk J Med Biol Res 1:5–12 35. Sener A, Malaisse WJ (1996) Hexose metabolism in pancreatic islets: apparent dissociation between the secretory and metabolic effects of D-fructose. Biochem Mol Med 59:182–186 36. Viñambres C, Villanueva-Peñacarrillo ML, Valverde I, Malaisse WJ (1997) Failure of D-fructose to stimulate protein biosynthesis in pancreatic islets. Biochem Mol Biol Int 41:571–574 37. Giroix M-H, Scruel O, Ladrière L, Sener A, Portha B, Malaisse WJ (1999) Metabolic and secretory interactions between D-glucose and D-fructose in islets from GK rats. Endocrinology 140:5556–5565 38. Scruel O, Giroix M-H, Sener A, Portha B, Malaisse WJ (1999) Metabolic and secretory response to D-fructose in islets from adult rats injected with streptozotocin during the neonatal period. Mol Genet Metab 68:86–90 39. Kyriazis GA, Soundarapandian MM, Tyrberg B (2012) Sweet taste receptor signaling in beta cells mediates fructose-induced potentiation of glucose-stimulated insulin secretion. Proc Natl Acad Sci U S A 109:E524–E532
