The pancreatic islets are richly innervated by autonomic nerves. The islet parasympathetic nerves emanate from intrapancreatic ganglia, which are controlled by preganglionic vagal nerves. The islet sympathetic nerves are postganglionic with the nerve cell bodies located in ganglia outside the pancreas. The sensory nerves originate from dorsal root ganglia near the spinal cord. Inside the islets, nerve terminals run close to the endocrine cells. In addition to the classic neurotransmitters acetylcholine and norepinephrine, several neuropeptides exist in the islet nerve terminals. These neuropeptides are vasoactive intestinal polypeptide, pituitary adenylate cyclase–activating polypeptide, gastrin-releasing polypeptide, and cocaine- and amphetamine-regulated transcript in parasympathetic nerves; neuropeptide Y and galanin in the sympathetic nerves; and calcitonin gene–related polypeptide in sensory nerves. Activation of the parasympathetic nerves and administration of their neurotransmitters stimulate insulin and glucagon secretion, whereas activation of the sympathetic nerves and administration of their neurotransmitters inhibit insulin but stimulate glucagon secretion. The autonomic nerves contribute to the cephalic phase of insulin secretion, to glucagon secretion during hypoglycemia, to pancreatic polypeptide secretion, and to the inhibition of insulin secretion, which is seen during stress. In rodent models of diabetes, the number of islet autonomic nerves is upregulated. This review focuses on neural regulation of islet function, with emphasis on the neuropeptides.
Since the discovery of nerves in the pancreatic islet by Paul Langerhans in his thesis from 1869, the neural-islet axis has been explored by a number of neuroanatomists, physiologists, and endocrinologists (rev. in 1–3). It is currently known that branches of the parasympathetic and sympathetic as well as the sensory nervous system innervate the islets with nerve terminals ending closely to the islet endocrine cells. It is also known that these nerves affect islet hormone secretion. The classic neurotransmitters in the islet autonomic nerves are acetylcholine and norepinephrine. During the last decades, the contribution to neural regulation of islet function also by neuropeptides has been established (rev. in 4). Several neuropeptides are localized to islet nerve terminals, are released from the pancreatic nerves upon nerve stimulation, and influence islet hormone secretion. It is also known that the islet innervation is altered in animal models of type 2 diabetes (5). The autonomic nervous system has also been suggested to be involved in the regulation of islet mass (6). The present review highlights the role of neuropeptides in islet function.
PARASYMPATHETIC NERVOUS SYSTEM
Anatomy and effects.
The parasympathetic nerves innervating the pancreatic islets emanate from the pancreatic ganglia, which are innervated by preganglionic parasympathetic nerves originating in the dorsal motor nucleus of the vagus. Activation of the parasympathetic nerves enhances insulin and glucagon secretion (1,3). These stimulatory effects are of physiological relevance under at least three conditions: 1) for the cephalic phase of insulin secretion during meal ingestion, 2) for the glucagon response to hypoglycemia, and 3) for pancreatic polypeptide (PP) secretion.
Parasympathetic nerves and cephalic phase of insulin secretion.
The cephalic phase of insulin secretion leads to the rapid and early increase in insulin levels during the first minutes after food ingestion (7). This is due to activation of olfactory-gustatory sensory receptors in association with psychological stimuli, which activates central parasympathetic nerves that stimulate insulin secretion. The importance of the cephalic phase of insulin secretion was recently examined by an approach to block the autonomic ganglia with the ganglionic blocker, trimetophane, in healthy subjects (7). In this approach, trimetophane was infused intravenously to inhibit the autonomic ganglia. A meal was served during the trimetophane infusion, and the trimetophane infusion was stopped 15 min later. It was first demonstrated that circulating insulin increased within the first 10 min after meal ingestion, which is before any rise in circulating glucose was observed. Of more importance in this context, however, was the second finding that this 10-min insulin response to meal ingestion was reduced by ∼75% by trimetophane (Fig. 1). This demonstrates that a neurally mediated cephalic phase of meal-related insulin secretion exists in humans. This reduction was accompanied by impairment of glucose elimination, even though the inhibited insulin response was limited to the first 15 min. Hence, the rapid and early insulin secretion contributes substantially to the glucose tolerance after meal ingestion. This may in turn be ascribed to the inhibition of hepatic glucose production by insulin.
Parasympathetic nerves and stimulation of glucagon secretion during hypoglycemia.
The parasympathetic nerves might also be of physiological importance for the stimulation of glucagon secretion during hypoglycemia (8). Glucagon secretion during counterregulation might be mediated by low glucose, low islet insulin, and high epinephrine, which all are consequences of hypoglycemia and all stimulate glucagon secretion. A possible contribution by the islet nerves was examined using the trimetophane protocol to inhibit autonomic ganglia in the presence of insulin-induced hypoglycemia in healthy subjects (9). It was found that the glucagon response to hypoglycemia (∼2.5 mmol/l glucose) was markedly reduced by trimetophane (Fig. 2). Hence, a significant proportion of the glucagon response to hypoglycemia depends on autonomic nerves.
Parasympathetic nerves and PP secretion.
The islet parasympathetic nerves seem of importance for PP secretion. This is evident from findings that vagus nerve stimulation increases PP secretion (10,11) and that the PP response to hypoglycemia is reduced by autonomic ganglionic blockade, as demonstrated in humans (9). In fact, the coupling of PP secretion to parasympathetic activity has suggested that plasma PP levels may be used as a marker or index of parasympathetic activity (12). However, the physiological relevance of the PP response to parasympathetic activation is not known.
Islet parasympathetic neurotransmitters.
It is well known that acetylcholine stimulates insulin secretion through a direct action on the islet β-cells (3). In addition, several neuropeptides are localized to islet parasympathetic nerve terminals and therefore potentially contribute to the islet effects of parasympathetic activation. These neuropeptides are vasoactive intestinal polypeptide (VIP), pituitary adenylate cyclase–activating polypeptide (PACAP), gastrin-releasing polypeptide (GRP), neuropeptide Y (NPY), and cocaine- and amphetamine-regulated transcript (CART) peptide (2–4,13–15). Figure 3 illustrates immunofluorescence of rodent pancreatic tissue showing the close proximity of these nerves with the pancreatic islets by showing VIP as an example: nerves co-harboring VIP and the cholinergic marker vesicular acetylcholine transporter (VAchT) enter the islets to terminate in close proximity to the islet endocrine cells (Figs. 3A–C). Figure 3 also shows nerves harboring GRP (Fig. 3G) and CART (Fig. 3H) in islets. Figure 3I–J shows that a great proportion of the CART-containing fibers are also VIP immunoreactive. Similar findings have previously been reported for PACAP (16) and, in the pig, for GRP (17) and, recently, for CART peptide (14,15). Therefore, VIP, PACAP, GRP, and CART are parasympathetic co-transmitter neuropeptides in autonomic nerve endings in the islets.
Effects of parasympathetic co-transmitters on insulin and glucagon secretion in mice.
VIP, PACAP, and GRP stimulate insulin and glucagon secretion when administered both in vivo in several species, including humans, and in vitro in isolated islets or the perfused pancreas (3,4,16–19). Here we report that VIP, PACAP, and GRP all potentiate glucose-stimulated insulin secretion (GSIS) and arginine-stimulated glucagon secretion, both when examined in isolated mouse islets (Figs. 4A and 5A) and when intravenously administered to mice together with glucose (Fig. 4C) or arginine (Fig. 5C). Increased insulin secretion has also been documented in transgenic mice overexpressing the VIP gene (20) or the PACAP gene (21) in the islet β-cells. The molecular basis of their effects is still far from fully understood, although it is established that VIP and PACAP signal through cAMP.
The importance of VIP and PACAP for islet physiology has been explored in model experiments using specific receptor antagonists and peptide ligand or receptor knockout mice. A study using PACAP−/− mice showed impaired GSIS in these mice (22). Also VIP−/− mice have been generated (23), but insulin secretion in these animals remains to be examined. Another strategy to explore the physiology of VIP and PACAP is to inhibit the activity or expression of their receptors. The two neuropeptides activate both VPAC1 and VPAC2 receptors and PACAP in addition to PAC1 receptors. Of these receptors, VPAC2 and PAC1 are expressed in islet cells (16). By disrupting the PAC1 receptor gene in mice, expression of truncated PAC1 receptors, which do not bind PACAP, evolves (24). PAC1R−/− mice display a marked reduction in the insulin response to both oral and intravenous glucose, showing that PACAP is of importance for a normal GSIS. Furthermore, the insulin response to intravenous administration of 2-deoxy-glucose (2-DG), is reduced in PAC1R−/− mice (Fig. 6). 2-deoxy-glucose competes with glucose for phosphorylation, which results in neuroglycopenia. This in turn activates the autonomic nerves, which results in a stimulation of insulin secretion after 2 min (25). Other studies have shown that PAC1 receptor antagonists pharmacologically inhibit the insulin response to oral glucose in mice (26) and to vagal nerve activation in the pig pancreas (27). A recent study compared the reduction in insulin secretion after gastric glucose versus intravenous glucose in PAC1R−/− mice by matching the glucose levels under these two conditions. The reduction was more marked after gastric glucose, again suggesting that PACAP contributes to the insulin response to oral glucose (28). The impairment of the insulin response in these mice suggests that PACAP may contribute to the cephalic phase of insulin secretion. In contrast, mice with VPAC2 gene deletion have a normal glucose tolerance during an oral glucose tolerance test in association with a reduced insulin response (29). This suggests increased insulin sensitivity in VPAC2−/− mice with an appropriate downregulation of the insulin response to maintain normal glucose tolerance, which indirectly would support normal β-cell function. A recent study has also demonstrated that PAC1R−/− mice display impaired glucagon secretion during hypoglycemia (30). Thus, PACAP may contribute to the parasympathetic involvement in the glucagon response to counterregulation, in addition to its potential physiological importance in regulating insulin secretion after 1) glucose stimulation, 2) vagal nerve activation, and 3) meal intake.
Because VIP and PACAP both strongly stimulate insulin secretion, they may be of potential interest in the treatment of type 2 diabetes. PACAP has been shown to reduce the hyperglycemia in rodent diabetes (high fat–fed mice and GK rats) (31). However, one problem is that PACAP stimulates glucagon secretion and the peptide has potent vasoactive effects, which would limit its usefulness in treatment. Instead, specific activation of VPAC2 receptors would be advantageous. Recently, a specific VPAC2 receptor agonist was described (32): it augments GSIS in both rodent and human islets and potentiates insulin secretion and glucose disposal in rats, thereby offering a novel target for treatment of type 2 diabetes based on islet neuropeptides.
The islet parasympathetic nerve endings also harbor GRP. GRP is released from the pancreas during parasympathetic nerve stimulation and stimulates insulin and glucagon secretion (3,4,17,19). It has been shown that the GRP receptor is expressed in islets (33), suggesting a direct action of the neuropeptide on islet cells. A study in GRP receptor gene–deficient mice has shown that the insulin response to oral glucose is impaired (34), which would support a role in the meal-related insulin response. It was also shown that insulin secretion in response to endogenous nerve activation in mice (by 2-deoxyglucose) is impaired in these mice, suggesting that GRP contributes to neurally mediated islet hormone secretion. These findings suggest that islet neuronal GRP, like VIP and PACAP, is of physiological importance.
It was recently also demonstrated that CART, an anorexigenic peptide that is highly expressed in the brain (rev. in 35), is also a neuropeptide of the rat and mouse pancreas (14,15, rev. in 36). Figure 3H–J shows the location of CART-containing nerves within an islet. Colocalization with VIP demonstrates the parasympathetic identity of the majority of the CART fibers. Recent studies have shown that CART affects islet hormone secretion (37). CART inhibits GSIS from isolated rat islets. On the other hand, CART potentiates GSIS augmented by glucagon-like peptide 1. Although there is hitherto no CART receptor identified, recent data suggest that CART exerts the potentiating effect on glucagon-like peptide 1–mediated GSIS via increased cAMP and the protein kinase A–dependent pathway (37). Furthermore, the potential impact of CART of islet function was demonstrated by using CART−/− mice (14). These mice displayed blunted GSIS both in vivo and in vitro, together with impaired glucose elimination.
SYMPATHETIC NERVOUS SYSTEM
Anatomy and function.
The islet sympathetic nerves are postganglionic, with their nerve cell bodies mainly located in the celiac ganglion or in the paravertebral sympathetic ganglia. Electrical stimulation of the sympathetic nerves inhibits insulin secretion and stimulates glucagon secretion (1–4). This is of physiological importance during stress and physical exercise to elicit a hyperglycemic response by increasing hepatic glucose delivery. An experimental tool for the study of function of the sympathetic nerves is administration of 6-hydroxydopamine to rodents. 6-hydroxydopamine is taken up by way of vesicular monoamine transporter localized to nerve endings of sympathetic neurons and selectively destroys sympathetic nerve terminals. As a consequence, islet nerves staining for the sympathetic marker tyrosine hydroxylase (TH) are absent in mice at 48 h after administration of 6-hydroxydopamine (38). This is associated with augmented GSIS (39,40), reduced insulin gene expression, and increased β-cell mass (38). These findings therefore suggest that the sympathetic nerves are of importance both for insulin secretion, insulin gene expression, and islet β-cell mass.
Islet sympathetic neurotransmitters.
The classic sympathetic neurotransmitter is norepinephrine, which inhibits GSIS and stimulates glucagon secretion (3,39). However, combined α- and β-adrenoceptor blockade does not prevent sympathetic nerve activation from inhibiting insulin secretion (41). This suggests that neurotransmitters other than norepinephrine contribute to some sympathetic islet effects. Neuropeptides localized to islet sympathetic nerve terminals are NPY and galanin (1–4). They are both released from the pancreas during sympathetic nerve activation (13,39) and, like sympathetic nerve stimulation, inhibit insulin secretion and stimulate glucagon secretion (3,4). Figure 3D–E illustrates that NPY nerves are localized to islets and the colocalization with TH illustrates the sympathetic nature of the NPY-containing nerve endings. It should be emphasized, however, that NPY is localized also to nerve endings that do not harbor TH. Also, galanin is localized to TH-containing nerve endings in the islets in a number of species (2,42). Both NPY and galanin are released from the pancreas during sympathetic nerve activation (13,43) and, like sympathetic nerve stimulation, inhibit insulin secretion and stimulate glucagon secretion (3,4). Figures 4B and D and 5B and D show that norepinephrine, NPY, and galanin all inhibit GSIS and augment arginine-stimulated glucagon secretion in isolated mouse islets and in vivo in mice. Because NPY and galanin thus mimic the effects of sympathetic nerve stimulation, they might contribute to sympathetic islet effects.
In the dog, galanin has been suggested to be of major importance in contributing to the inhibitory influence of sympathetic nerve activation on insulin secretion (43). Also in mice, galanin seems to be of physiological impact because immunoneutralization of galanin prevents the inhibition of insulin secretion, which is seen during a stress model (swimming) in mice (44). To study the physiological impact of galanin on islet function in more detail, mice with a loss-of-function mutation in the galanin gene have been examined (45). The inhibition of insulin secretion that is seen after administration of 2-deoxyglucose (and that reflects activation of sympathetic nerves) was impaired in galanin−/− mice. This further supports a contribution by galanin of the response to sympathetic activation.
SENSORY NERVES
The islets are innervated by sensory nerves that harbor calcitonin gene–related polypeptide (CGRP) (1–4). The fibers leave the pancreas along the sympathetic fibers with the splanchnic nerves and reach the spinal cord. Figure 3H illustrates an islet fiber harboring CGRP. The relevance of these nerves for islet function is far from understood. These sensory nerves may be targeted by the use of the toxin capsaicin, which causes degeneration of small unmyelinated C-fibers. Capsaicin is an agonist for the transient receptor potential vanilloid receptor (TRPV1); acutely, it activates the receptors, which leads to a release of the neurotransmitters. After the acute effect, however, capsaicin leads to loss of unmyelinated sensory fibers in conjunction with a substantial number of CGRP nerves (46). If capsaicin is given to neonatal rodents, the sensory deactivation is permanent, whereas if given to adults, the deactivation is transient. One effect of capsaicin-induced sensory deactivation is increased insulin secretion (47). This would suggest that sensory nerve activation in islets inhibits insulin secretion.
Islet sensory neurotransmitters.
CGRP nerves are scattered through the pancreas but with particular density around small blood vessels and islets (48). Furthermore, exogenous administration of CGRP inhibits GSIS (48). This suggests that sensory nerves inhibit insulin secretion; however, the physiological importance of this remains elusive. It should be mentioned in this context that CART is also present in CGRP-containing fibers in rat and mouse pancreas (14,15). The biological significance of CART in these fibers needs further investigation.
Sensory deafferentation and treatment of diabetes.
Since sensory nerves apparently inhibit insulin secretion, possibly through CGRP, and since sensory deafferentation by capsaicin increases insulin secretion, it has been proposed that sensory deactivation might be a novel target for treatment of type 2 diabetes. This hypothesis was substantiated in a recent study in obese Zucker rats (49). This novel potential neurally based therapeutic approach has been further explored by using the toxin resiniferatoxin. This is a vanilloid that has the same mechanism in causing deactivation/degeneration of C-fibers (and Aδ-fibers) as capsaicin but is less toxic. By administering resiniferatoxin to obese Zucker rats (50) and to Zucker diabetic fatty (ZDF) rats (51), improved glycemia has been observed, together with improved insulin secretion, as demonstrated in ZDF rats.
ISLET NERVES AND DIABETES
Several studies have indicated altered islet neurohormonal influences in models of insulin resistance and type 2 diabetes, and therefore it has been speculated whether islet nerve dysfunction may contribute to the development of type 2 diabetes. A study in high fat–fed rats, which is a model of glucose intolerance and type 2 diabetes, disclosed an increased islet innervation (52). Furthermore, insulin resistance in high fat–fed mice is accompanied by augmented insulin secretion after cholinergic activation as a sign of cholinergic hypersensitivity (53) and as a sign that the hyperinsulinemia in ob/ob mice is highly sensitive to atropine (54). Indeed, cholinergic activation by carbachol normalizes insulin secretion in high fat–fed mice (55). It was therefore of interest that reduced islet innervation was evident in a model of type 2 diabetes, the Chinese hamster (5).
We present here immunocytochemical data of islets from two rodent models of diabetes. The models are the Goto-Kakizaki (GK) rats and the db/db mice. The GK rat model is one of the best-described models of type 2 diabetes, as recently was reviewed (56). Its basis is a β-cell defect, which occurs through changes in several independent genes, leading to impaired insulin secretion in combination with metabolic impairment reducing β-cell neogenesis and a potential acquired loss of β-cell differentiation due to glucotoxicity. The db/db mouse is characterized by a leptin receptor deficiency, resulting in insulin resistance and impaired islet function (57). Figures 7 and 8 show that in both these two rodent models of diabetes, the islet innervation is upregulated. This emphasizes that islet innervation is of importance for the islet dysfunction in diabetes. We also found that certain neuropeptides, e.g., VIP and CART (Table 1) (37), are expressed in islet endocrine cells in the diabetic rodents. This is similar to our previous observation of marked expression of NPY in islet β-cells in dexamethasone-induced insulin resistance in rats (58). We propose that augmented endocrine cell expression of neurotransmitters in the islets is a sign of islet adaptation for normalization of glucose homeostasis. We suggest the term “neuro-islet plasticity” for this phenomenon; a thorough examination of which needs to be undertaken now.
A novel potential effect of parasympathetic nerves, which may be of pathophysiological relevance for type 2 diabetes, is stimulation of β-cell mass. Regulation of β-cell mass has recently come into focus because of studies suggesting that β-cell mass is reduced in type 2 diabetes (6,59). In adult life, β-cell growth and renewal is mainly regulated by replication and differentiation from islet precursor cells within islets and within the duct epithelium (60). An important mechanism for increase in β-cell mass may therefore be the increased demand created by insulin resistance; therefore, islets are enlarged in, for instance, the pre-diabetic state and obesity. On the other hand, in overt type 2 diabetes, β-cell mass seems to be reduced, which may be ascribed to exaggerated apoptosis triggered by, for instance, glucotoxicity or lipotoxicity. However, neural effects may also contribute to these perturbations. Thus, it has been demonstrated that vagotomy attenuates the islet hyperplasia in ob/ob mice, the mechanism of which needs to be further explored (61). From a therapeutic point of view, the potential of the incretin hormone glucagon-like peptide 1 to augment islet β-cell mass, presumably by both increasing neogenesis and inhibiting apoptosis, has been extensively discussed (60). However, islet neuropeptides may also be of interest in this context; for example, exogenous administration of PACAP has been shown to maintain β-cell mass in an experimental diabetes model in mice, which is due to increased β-cell apoptosis (β-cell overexpression of calmodulin [62]).
CONCLUSIONS AND OUTSTANDING ISSUES
The pancreatic islets are innervated by the autonomic nervous system, which affect islet function through both classic neurotransmitters and neuropeptides. Table 1 summarizes the neuropeptides, which are of greatest importance in this context. The islet nerves seem of physiological relevance for 1) the insulin secretion during the cephalic phase of meal ingestion, 2) the glucagon response to hypoglycemia, 3) the PP release, and 4) the islet stress response. The islet nerves may also be of significance for the regulation of islet mass. Novel potential targets for treatment of diabetes may evolve from studies of islet effects of autonomic nerves and neuropeptides. Of most interest at the moment is stimulation of insulin secretion by 1) cholinergic agonism, 2) VPAC2 receptor agonists, 3) PACAP, 4) CART in combination with glucagon-like peptide 1, and 5) sensory nerve deactivation. However, much remains to be established regarding the potential role for the autonomic nerves in relation to islet function. Complex issues relate to mechanisms of activation of the autonomic nerves and regulation of release of the neurotransmitters in relation to each other and the possibility of synergistic action of the neurotransmitters. Other issues relate to the involvement of neuropeptides in islet physiology in humans as well as to potential importance of the autonomic nerves for the islet compensation to insulin resistance, to diabetes development, and as targets for treatment of diabetes. Therefore, although almost 140 years have passed since the first description of islet nerves by Paul Langerhans, much effort has yet to be paid to understand the role for these nerves in islet physiology and pathophysiology.
Serum insulin and plasma glucose levels in six healthy women subjected to a 30-min intravenous infusion of the ganglionic antagonist trimetophane or saline (from −15 to +15 min). At time 0, a standardized mixed meal was served. Small insert in upper panel shows the insulin levels between 0 and 10 min in detail. Data are means ± SE. Asterisks indicate probability level of random difference between the two protocols, as revealed by ANOVA: *P < 0.05, **P < 0.01. Reproduced with permission from the American Diabetes Association (7).
Serum insulin and plasma glucose levels in six healthy women subjected to a 30-min intravenous infusion of the ganglionic antagonist trimetophane or saline (from −15 to +15 min). At time 0, a standardized mixed meal was served. Small insert in upper panel shows the insulin levels between 0 and 10 min in detail. Data are means ± SE. Asterisks indicate probability level of random difference between the two protocols, as revealed by ANOVA: *P < 0.05, **P < 0.01. Reproduced with permission from the American Diabetes Association (7).
Plasma glucagon in healthy postmenopausal women before and during insulin and glucose infusions producing hypoglycemia of ∼2.5 mmol/l during intravenous infusion of the ganglionic antagonist trimetophane (○) or saline (•). Data are means ± SE. Adapted with permission from the American Diabetes Association (9).
Plasma glucagon in healthy postmenopausal women before and during insulin and glucose infusions producing hypoglycemia of ∼2.5 mmol/l during intravenous infusion of the ganglionic antagonist trimetophane (○) or saline (•). Data are means ± SE. Adapted with permission from the American Diabetes Association (9).
Immunofluorescence micrographs of rodent islets. A–C: Mouse islet double stained for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note the high degree of colocalization, illustrating that VIP immunoreactive fibers are cholinergic. D–F: Part of rat islet double stained for NPY (D) and TH (E), merged in F. A great proportion of the NPY immunoreactive fibers are adrenergic. G: Bombesin/GRP immunoreactive fiber in a mouse islet. H: CGRP immunoreactive fiber in a mouse islet. I–K: Part of rat islet double stained for VIP (I) and CART (J), merged in K. A great proportion of the CART-containing fibers are also VIP immunoreactive. Note that CART immunoreactivity is also seen in a δ-cell (J). In brief, pancreatic tissue from normal C57BL/6J and normal Sprague-Dawley rats were fixed overnight in Stefanini’s solution and, after being frozen on dry ice, were cut in a cryostat, mounted on slides, and incubated with a guinea pig anti-VIP antibody (code 8701, dilution 1:1,280; EuroDiagnostica, Malmö, Sweden), rabbit anti-TH (code P401010-0, dilution 1:320; Pel-Freeze Biologicals, Rogers, AR), rabbit anti–COOH-terminal flanking peptide of NPY (dilution 1:1,280; Genosys Biotech, Pampisford, U.K.), rabbit anti-CART (code 12/D, dilution 1:1,280; Cocalico, Reamstown, PA), guinea pig anti-CGRP (code M8513, dilution 1:640; EuroDiagnostica), and anti-bombesin/GRP (code 81051, dilution 1:600; EuroDiagnostica). Sections were incubated for 1 h at room temperature with a secondary antibody, coupled with fluorescein isothiocyanate or Texas red with specificity for IgG (dilution 1:100 and 1:400, respectively; DAKO, Copenhagen, Denmark).
Immunofluorescence micrographs of rodent islets. A–C: Mouse islet double stained for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note the high degree of colocalization, illustrating that VIP immunoreactive fibers are cholinergic. D–F: Part of rat islet double stained for NPY (D) and TH (E), merged in F. A great proportion of the NPY immunoreactive fibers are adrenergic. G: Bombesin/GRP immunoreactive fiber in a mouse islet. H: CGRP immunoreactive fiber in a mouse islet. I–K: Part of rat islet double stained for VIP (I) and CART (J), merged in K. A great proportion of the CART-containing fibers are also VIP immunoreactive. Note that CART immunoreactivity is also seen in a δ-cell (J). In brief, pancreatic tissue from normal C57BL/6J and normal Sprague-Dawley rats were fixed overnight in Stefanini’s solution and, after being frozen on dry ice, were cut in a cryostat, mounted on slides, and incubated with a guinea pig anti-VIP antibody (code 8701, dilution 1:1,280; EuroDiagnostica, Malmö, Sweden), rabbit anti-TH (code P401010-0, dilution 1:320; Pel-Freeze Biologicals, Rogers, AR), rabbit anti–COOH-terminal flanking peptide of NPY (dilution 1:1,280; Genosys Biotech, Pampisford, U.K.), rabbit anti-CART (code 12/D, dilution 1:1,280; Cocalico, Reamstown, PA), guinea pig anti-CGRP (code M8513, dilution 1:640; EuroDiagnostica), and anti-bombesin/GRP (code 81051, dilution 1:600; EuroDiagnostica). Sections were incubated for 1 h at room temperature with a secondary antibody, coupled with fluorescein isothiocyanate or Texas red with specificity for IgG (dilution 1:100 and 1:400, respectively; DAKO, Copenhagen, Denmark).
Effects of islet neurotransmitters on insulin secretion in vivo and in vitro in NMRI mice. A and B: Insulin secretion from overnight-incubated collagenase-isolated islets after incubation for 60 min in a medium containing various glucose concentrations with or without addition of islet cholinergic co-transmitters (Fig. 4A) or islet adrenergic co-transmitters (Fig. 4B). Synthetic porcine VIP (100 nmol/l), ovine PACAP-38 (100 nmol/l), porcine GRP (100 nmol/l), norepinephrine (1 μmol/l), porcine NPY (100 nmol/l), or rat galanin (100 nmol/l) were used. There were 24–32 incubations at each data point; means are shown. C and D: Plasma insulin levels after intravenous administration of glucose (1 g/kg) alone or together with synthetic porcine VIP (1.5 nmol/kg), ovine PACAP-38 (1.3 nmol/kg), porcine GRP (4 nmol/kg), norepinephrine (320 nmol/kg), porcine NPY (1.1 nmol/kg), or rat galanin (2 nmol/kg) in anesthetized mice. There were 16–20 animals in each group. Means are shown. For the in vitro studies, pancreatic islets from normal NMRI mice were isolated by collagenase digestion and handpicked under microscope. After overnight incubation in RPMI medium, the islets were preincubated in HEPES balanced salt solution (HBSS) containing 125 mmol/l NaCl, 5.9 mmol/l KCl, 1.28 mmol/l CaCl2, 1.2 mmol/l MgCl2, 25 mmol/l HEPES (pH 7.4), 3.3 mmol/l glucose, and 0.1% fatty acid–free bovine albumin (Boehringer Mannheim, Mannheim, Germany) for 60 min. Islets in groups of three were then incubated in 200 μl HBSS for 60 min at 37°C, with varying concentrations of glucose alone or with addition of the peptides (all peptides from Peninsula Laboratories Europe, Merseyside, U.K.). For the in vivo studies, plasma insulin levels were determined after intravenous administration of glucose (1 g/kg) alone or together with the peptides in NMRI mice anesthetized with 0.5 mg/mouse fluanison, 0.02 mg/mouse fentanyl (Hypnorm; Janssen, Beerse, Belgium), and 0.25 mg/mouse midazolam (Dormicum; Hoffman-LaRoche, Basel, Switzerland). Samples (75 μl) were taken from the retrobulbar intraorbital capillary plexus before and at various time points after the intravenous injections. Plasma was separated and analyzed for insulin with radioimmunoassay (Linco Research, St. Charles, MO).
Effects of islet neurotransmitters on insulin secretion in vivo and in vitro in NMRI mice. A and B: Insulin secretion from overnight-incubated collagenase-isolated islets after incubation for 60 min in a medium containing various glucose concentrations with or without addition of islet cholinergic co-transmitters (Fig. 4A) or islet adrenergic co-transmitters (Fig. 4B). Synthetic porcine VIP (100 nmol/l), ovine PACAP-38 (100 nmol/l), porcine GRP (100 nmol/l), norepinephrine (1 μmol/l), porcine NPY (100 nmol/l), or rat galanin (100 nmol/l) were used. There were 24–32 incubations at each data point; means are shown. C and D: Plasma insulin levels after intravenous administration of glucose (1 g/kg) alone or together with synthetic porcine VIP (1.5 nmol/kg), ovine PACAP-38 (1.3 nmol/kg), porcine GRP (4 nmol/kg), norepinephrine (320 nmol/kg), porcine NPY (1.1 nmol/kg), or rat galanin (2 nmol/kg) in anesthetized mice. There were 16–20 animals in each group. Means are shown. For the in vitro studies, pancreatic islets from normal NMRI mice were isolated by collagenase digestion and handpicked under microscope. After overnight incubation in RPMI medium, the islets were preincubated in HEPES balanced salt solution (HBSS) containing 125 mmol/l NaCl, 5.9 mmol/l KCl, 1.28 mmol/l CaCl2, 1.2 mmol/l MgCl2, 25 mmol/l HEPES (pH 7.4), 3.3 mmol/l glucose, and 0.1% fatty acid–free bovine albumin (Boehringer Mannheim, Mannheim, Germany) for 60 min. Islets in groups of three were then incubated in 200 μl HBSS for 60 min at 37°C, with varying concentrations of glucose alone or with addition of the peptides (all peptides from Peninsula Laboratories Europe, Merseyside, U.K.). For the in vivo studies, plasma insulin levels were determined after intravenous administration of glucose (1 g/kg) alone or together with the peptides in NMRI mice anesthetized with 0.5 mg/mouse fluanison, 0.02 mg/mouse fentanyl (Hypnorm; Janssen, Beerse, Belgium), and 0.25 mg/mouse midazolam (Dormicum; Hoffman-LaRoche, Basel, Switzerland). Samples (75 μl) were taken from the retrobulbar intraorbital capillary plexus before and at various time points after the intravenous injections. Plasma was separated and analyzed for insulin with radioimmunoassay (Linco Research, St. Charles, MO).
Effects of islet neurotransmitters on glucagon secretion in vivo and in vitro in NMRI mice. A and B: Glucagon secretion from overnight incubated collagenase-isolated islets after incubation for 60 min in a medium containing glucose at 5 mmol/l and arginine at 20 mmol/l with or without addition of islet cholinergic co-transmitters (Fig. 5A) or islet adrenergic co-transmitters (Fig. 5B). Synthetic porcine VIP (100 nmol/l), ovine PACAP-38 (100 nmol/l), porcine GRP (100 nmol/l), norepinephrine (1 μmol/l), porcine NPY (100 nmol/l), or rat galanin (100 nmol/l) were used. There were 16 incubations at each data point; means are shown. C and D: Plasma glucagon levels after intravenous administration of arginine (0.25 g/kg) alone or together with synthetic porcine VIP (1.5 nmol/kg), ovine PACAP-38 (1.3 nmol/kg), porcine GRP (4 nmol/kg), norepinephrine (320 nmol/kg), porcine NPY (1.1 nmol/kg), or rat galanin (2 nmol/kg) in anesthetized mice. There were eight animals in each group. Means are shown. Symbols are same as in Fig. 4. Glucagon was determined with radioimmunoassay (Linco Research).
Effects of islet neurotransmitters on glucagon secretion in vivo and in vitro in NMRI mice. A and B: Glucagon secretion from overnight incubated collagenase-isolated islets after incubation for 60 min in a medium containing glucose at 5 mmol/l and arginine at 20 mmol/l with or without addition of islet cholinergic co-transmitters (Fig. 5A) or islet adrenergic co-transmitters (Fig. 5B). Synthetic porcine VIP (100 nmol/l), ovine PACAP-38 (100 nmol/l), porcine GRP (100 nmol/l), norepinephrine (1 μmol/l), porcine NPY (100 nmol/l), or rat galanin (100 nmol/l) were used. There were 16 incubations at each data point; means are shown. C and D: Plasma glucagon levels after intravenous administration of arginine (0.25 g/kg) alone or together with synthetic porcine VIP (1.5 nmol/kg), ovine PACAP-38 (1.3 nmol/kg), porcine GRP (4 nmol/kg), norepinephrine (320 nmol/kg), porcine NPY (1.1 nmol/kg), or rat galanin (2 nmol/kg) in anesthetized mice. There were eight animals in each group. Means are shown. Symbols are same as in Fig. 4. Glucagon was determined with radioimmunoassay (Linco Research).
Plasma insulin response to intravenous administration of 2-deoxyglucose (2-DG; 0.5 g/kg) in anesthetized PAC1R−/− mice (○) or their wild-type counterparts (•). Asterisks indicate probability level of random difference between the groups: *P < 0.05, **P < 0.01, ***P < 0.001.
Plasma insulin response to intravenous administration of 2-deoxyglucose (2-DG; 0.5 g/kg) in anesthetized PAC1R−/− mice (○) or their wild-type counterparts (•). Asterisks indicate probability level of random difference between the groups: *P < 0.05, **P < 0.01, ***P < 0.001.
Immunofluorescence micrographs of GK rat islets illustrating that the numbers of both parasympathetic and sympathetic nerves are greatly increased in the islets. A–C: Double staining for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note also the high degree of colocalization (exemplified by arrow heads). D–F: Double staining for NPY (D) and TH (E), merged in F. NPY is largely colocalized with TH (exemplified by arrow heads); note that a substantial population of the NPY immunoreactive varicosities are devoid of TH, presumably representing parasympathetic fibers.
Immunofluorescence micrographs of GK rat islets illustrating that the numbers of both parasympathetic and sympathetic nerves are greatly increased in the islets. A–C: Double staining for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note also the high degree of colocalization (exemplified by arrow heads). D–F: Double staining for NPY (D) and TH (E), merged in F. NPY is largely colocalized with TH (exemplified by arrow heads); note that a substantial population of the NPY immunoreactive varicosities are devoid of TH, presumably representing parasympathetic fibers.
Immunofluorescence micrographs of db/db mouse islets illustrating that both parasympathetic and sympathetic nerves are greatly increased in the islets. A–C: Double staining for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note the high degree of colocalization (exemplified by arrow heads). D: Staining for NPY. E: Staining for TH. F: Staining for CART.
Immunofluorescence micrographs of db/db mouse islets illustrating that both parasympathetic and sympathetic nerves are greatly increased in the islets. A–C: Double staining for VIP (A) and vesicular acetylcholine transporter (B), merged in C. Note the high degree of colocalization (exemplified by arrow heads). D: Staining for NPY. E: Staining for TH. F: Staining for CART.
Neuropeptides that are localized to islet autonomic nerves: their primary and main localization in normal rodent islets, in GK rats, and in ob/ob mice and their effect on insulin and glucagon secretion
Neuropeptide . | Localization . | . | . | Effects . | . | |||
---|---|---|---|---|---|---|---|---|
. | Normal rodents . | GK rat . | db/db mouse . | Insulin secretion . | Glucagon secretion . | |||
VIP | PS nerves | Innervation ↑/β-Cells | Innervation ↑/β-Cells | Stimulation | Stimulation | |||
PACAP | PS nerves | NT | NT | Stimulation | Stimulation | |||
GRP | PS nerves | NT | NT | Stimulation | Stimulation | |||
NPY | S nerves | Innervation ↑ | Innervation ↑ | Inhibition | Stimulation | |||
Galanin | S nerves | NT | NT | Inhibition | Stimulation | |||
CGRP | Sensory nerves/ D-cells | Innervation ↑ | Unchanged | Inhibition | Stimulation | |||
CART | PS nerves | Innervation ↑/β-Cells | Innervation ↑/β-Cells | Stimulation | Inhibition |
Neuropeptide . | Localization . | . | . | Effects . | . | |||
---|---|---|---|---|---|---|---|---|
. | Normal rodents . | GK rat . | db/db mouse . | Insulin secretion . | Glucagon secretion . | |||
VIP | PS nerves | Innervation ↑/β-Cells | Innervation ↑/β-Cells | Stimulation | Stimulation | |||
PACAP | PS nerves | NT | NT | Stimulation | Stimulation | |||
GRP | PS nerves | NT | NT | Stimulation | Stimulation | |||
NPY | S nerves | Innervation ↑ | Innervation ↑ | Inhibition | Stimulation | |||
Galanin | S nerves | NT | NT | Inhibition | Stimulation | |||
CGRP | Sensory nerves/ D-cells | Innervation ↑ | Unchanged | Inhibition | Stimulation | |||
CART | PS nerves | Innervation ↑/β-Cells | Innervation ↑/β-Cells | Stimulation | Inhibition |
NT, not tested.
This article is based on a presentation at a symposium. The symposium and the publication of this article were made possible by an unrestricted educational grant from Servier.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Article Information
These studies have been supported by the Swedish Research Council (grants 6834 and 4499), Swedish Diabetes Association, Albert Påhlson Foundation, The Royal Physiographic Society, Novo Nordisk Foundation, Tore Nilsson, Åke Wiberg and Gyllenstiernska-Krapperup Foundations, Region Skåne, and the Faculty of Medicine, Lund University.