This Article Abstract Full Text (PDF) All Versions of this Article: 28/1/84    most recent Final Manuscript Author Manuscript Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gromada, J. Articles by Wollheim, C. B. Search for Related Content PubMed PubMed Citation Articles by Gromada, J. Articles by Wollheim, C. B.

-Cells of the Endocrine Pancreas: 35 Years of Research but the Enigma Remains

Novartis Institutes for BioMedical Research (J.G.), Cambridge, Massachusetts 02139; and Department of Cell Physiology and Metabolism (I.F., C.B.W.), University Medical Centre, 1211 Geneva 4, Switzerland

Correspondence: Address all correspondence and requests for reprints to: Jesper Gromada, Novartis Institutes for BioMedical Research, 100 Technology Square, Cambridge, Massachusetts 02139. E-mail: jesper.gromada{at}novartis.com

	Abstract

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
Glucagon, a hormone secreted from the -cells of the endocrine pancreas, is critical for blood glucose homeostasis. It is the major counterpart to insulin and is released during hypoglycemia to induce hepatic glucose output. The control of glucagon secretion is multifactorial and involves direct effects of nutrients on -cell stimulus-secretion coupling as well as paracrine regulation by insulin and zinc and other factors secreted from neighboring ß- and -cells within the islet of Langerhans. Glucagon secretion is also regulated by circulating hormones and the autonomic nervous system. In this review, we describe the components of the -cell stimulus secretion coupling and how nutrient metabolism in the -cell leads to changes in glucagon secretion. The islet cell composition and organization are described in different species and serve as a basis for understanding how the numerous paracrine, hormonal, and nervous signals fine-tune glucagon secretion under different physiological conditions. We also highlight the pathophysiology of the -cell and how hyperglucagonemia represents an important component of the metabolic abnormalities associated with diabetes mellitus. Therapeutic inhibition of glucagon action in patients with type 2 diabetes remains an exciting prospect.

I. Introduction

II. Islet Endocrine Cell Composition

A. Islet microcirculation

B. Junctional communication between -cells

C. Islet innervation

III. Paracrine, Autocrine, and Hormonal Regulation of Glucagon Secretion

A. Insulin

B. Zinc

C. GABA

D. Glutamate

E. Somatostatin

F. Ghrelin

G. GLP-1

H. Glucagon

IV. Autonomic Regulation of Glucagon Secretion

V. Transcriptional Control of Pancreatic -Cell Development

VI. The Glucagon Gene: Transcriptional Control and Proglucagon Processing

VII. -Cell Stimulus Secretion Coupling

A. Ion channels present in the -cell plasma membrane

B. Regulation of electrical activity in rat -cells

C. Regulation of electrical activity in mouse -cells

D. Metabolism of the -cell

E. Intracellular Ca²⁺ homeostasis

F. Regulation of exocytosis of glucagon-containing granules

G. Pharmacology

VIII. -Cell Pathophysiology and the Treatment of Diabetes

IX. Summary and Conclusions

	I. Introduction

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
THE HISTORY OF glucagon begins with that of insulin. In 1921, when F. Banting and C. Best tested their first pancreatic extracts in depancreatized dogs, they observed that insulin-induced hypoglycemia was preceded by a transient, rather mild hyperglycemia, and they thought that this unwanted effect was due to epinephrine release (1). Murlin et al. (2) must be credited with the discovery of glucagon in 1923, because they suggested that the early hyperglycemic effect of the pancreatic extracts was due to a contaminant with glucogenic properties that they also proposed to call "glucagon," or the mobilizer of glucose. In a classical paper published in 1948, Sutherland and de Duve (3) established the -cells of the pancreas as being the source of glucagon. At about the same time, Foá et al. (4, 5, 6) conducted elegant cross-circulation experiments in anesthetized dogs and suggested that hypoglycemia was triggering the release of glucagon by the pancreas. The description by Unger et al. (7, 8, 9) between 1959 and 1962 of a glucagon RIA made it possible to investigate the physiology of glucagon and its role in various diseases and disorders. Glucagon is a sensitive and timely regulator of glucose homeostasis in vivo in both animals and humans. Small doses of glucagon are sufficient to induce rapid but transient glucose elevations consistent with its role as a counterregulatory hormone (10, 11, 12). To increase blood glucose, glucagon promotes hepatic glucose output by stimulating glycogenolysis and gluconeogenesis and by decreasing glycogenesis and glycolysis in a concerted fashion via multiple mechanisms.

The physiological defenses against falling plasma glucose concentrations include decreased pancreatic ß-cell insulin secretion as well as increased glucagon and adrenomedullary epinephrine secretion (13) (Fig. 1). Under normal physiological conditions, the intraislet paracrine and endocrine interactions, especially reduced intraislet insulin and zinc release from the pancreatic ß-cells, promote glucagon secretion. A direct stimulatory effect of low glucose on the -cell seems of little physiological importance and, at least for the rat -cell, high glucose augments glucagon release (14, 15). In addition to peripheral glucose sensing, an essential role of glucose-responsive neurons in the ventromedial hypothalamus (VMH) has been established for the regulation of glucagon secretion and for glucose homeostasis (16, 17, 18). All of these defense mechanisms are compromised in type 1 diabetes and advanced type 2 diabetes. This involves absent insulin response resulting from ß-cell failure and loss of glucagon response. This is probably caused by loss of the decrement in intraislet insulin and/or zinc that normally results in enhanced glucagon secretion. In the setting of absent insulin and glucagon responses, attenuated epinephrine responses cause the clinical syndrome of defective glucose counterregulation that is associated with a much greater risk of severe hypoglycemia (19).

View larger version (37K): [in this window] [in a new window]   FIG. 1. Physiological defenses against hypoglycemia. It is important to note that the decreased insulin secretion and increased glucagon release are lost and the increase in epinephrine is often attenuated in type 1 and advanced type 2 diabetes. AA, Amino acid.

	II. Islet Endocrine Cell Composition

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
Pancreatic -cells were discovered in 1907 as histologically distinct cells from the ß-cells of the islet of Langerhans (20). The -cells are one of four distinct polypeptide-secreting islet cell types: glucagon-secreting -cells, insulin-producing ß-cells, somatostatin-releasing -cells, and pancreatic polypeptide (PP)-secreting cells. Recently, ghrelin-producing cells have also been observed in the islet (21, 22, 23). A recent study has shown that ß-, -, and -cells are scattered throughout the human islet (24). Thus, human islets do not show the anatomical subdivisions like rodent islets where the ß-cells are concentrated in the core of the islet, and - and -cells are located in the mantle (24). The cytoarchitecture of the human islet, where most of the ß-cells (71%) showed associations with other endocrine cells, suggest unique paracrine interactions. Most of the ß-, -, and -cells in human islets were aligned along blood vessels with no particular order or arrangement, suggesting that islet microcirculation likely does not determine the order of paracrine interactions (24). This would contrast with the situation in the rat islet (25) but needs to be demonstrated directly. In type 1 diabetic patients, where ß-cells are lost, -cells comprise approximately 75% of the total cell number (26), although the absolute mass (and therefore the ratio) of the -, -, and PP cells does not appear to be altered (27). However, in type 2 diabetes, -cell hyperplasia occurs (27), whereas the ß-cell mass is probably reduced as a result of increased apoptosis (28). Early studies also associated an increase in -cell number with age (29).

A. Islet microcirculation
Anatomical studies have shown that islets are densely vascularized, with at least one arteriole supplying every islet (30). This would permit simultaneous exposure of islets to changes in arterial milieu and rate of flow (31). In vivo studies in rat using fluorescent microscopy to monitor the flow of microspheres or albumin in a single islet located in the head of the pancreas indicated that the blood supplied by the arteriole flows first into capillaries located in one pole of the islet mantle, then traverses the islet core either directly or via the mantle to the opposite pole of the islet, where it exits via venules (32). Such a direction of flow would provide for paracrine actions of both - and -cell secretory products on downstream ß-cells and subsequent effects of ß-cell secretory products on cells located in the mantle of the venular pole. However, this possibility is not supported by physiological studies in the isolated pancreata of rats and humans, where perfusion with antisomatostatin antibodies (somatostatin is an inhibitor of both insulin and glucagon secretion) in the anterograde direction had no effect on insulin or glucagon secretion (33, 34, 35). In contrast, an increase in both glucagon and insulin secretion was observed when the antibody was perfused through the pancreata in the retrograde direction, suggesting that the -cells are in fact downstream of both - and ß-cells. This was supported by early anatomical studies in the rat islet showing arterial supply first to the ß-cell enriched core (25). In a separate study, perfusion of the human pancreas with an antibody against somatostatin in the anterograde direction had a very small but significant stimulatory effect on insulin secretion, and also glucagon secretion, the latter in low glucose conditions only, in support of the pole-to-pole direction of blood flow (36). A thorough anatomical examination of islet microvascular flow in different regions of the pancreas is now timely. This is particularly pertinent in the case of the mouse, where information is severely lacking, and may provide answers to the conflicting findings from the early anatomical vs. physiological studies.

B. Junctional communication between -cells
We have noted that, when measured as a percentage of content, the amount of glucagon released from fluorescence-activated cell sorted (FACS) -cells is much greater than that from intact islets (7.6 ± 0.9%, n = 9, 30-min incubation; vs. 0.53 ± 0.1%, n = 7, 1-h incubation, respectively, in the presence of 2.5 mM glucose). A high rate of basal glucagon secretion was also observed from dispersed (unsorted) islet cells (6.1 ± 1.1% of content, n = 4, 30-min incubation), indicating that the loss of intercellular contacts underlies the increased rate of basal secretion, rather than the absence of or reduction in exposure to paracrine inhibitory factors (I. Franklin and C. B. Wollheim, unpublished data). In isolated -cells, the increased basal secretion rate was not reduced by the removal of extracellular Ca²⁺ (14). To investigate whether junctional communication exists between neighboring -cells and whether these contacts are indeed required for normal rates of basal glucagon release, as previously observed for insulin secretion in ß-cells (37, 38), we attempted to reform contacts between cells by reaggregating the isolated -cell fraction. Reaggregating the cells caused a 10-fold reduction in the rate of glucagon release in basal glucose conditions, without altering the response to the secretagogue pyruvate, indicating that intercellular contacts may be necessary for normal basal secretion. Doubling the density of cells per well from 20,000 to 40,000 did not alter the rate of basal glucagon release, demonstrating that the altered rate of glucagon release from the reaggregated cells was not due to an increase in the local concentration of secreted autocrine factors. Immunocytochemical analysis of the reaggregated cell fraction did not reveal positive staining for either of the gap junction proteins connexin 36 or connexin 43 (I. Franklin and C. B. Wollheim, unpublished data), although the involvement of other members of the connexin family in the formation of gap junctions between rat -cells cannot be excluded. Connexin expression has been detected in FACS-isolated non-ß-cells from rats and mice (39, 40). Gap junctions, composed of connexin 36, are required for normal glucose-induced insulin secretion by rat ß-cells (38, 41). An understanding of the mechanisms regulating basal rates of glucagon secretion may lead to the identification of novel drug targets for controlling hyperglucagonemia in diabetes.

C. Islet innervation
Pancreatic islets are richly innervated to enable autonomic regulation of endocrine cell hormone secretion. The most extensively studied are the sympathetic (adrenergic) and parasympathetic (cholinergic) nerves, which can project deeply into the islet, but other types of sensory neurons have also been detected, including GABAergic nerve bodies (42). The sympathetic nervous system is activated by hypoglycemia or exercise stress, causing the release of norepinephrine and other neurotransmitters as well as neuropeptides into the islet (for review, see Ref. 43). These agents trigger glucagon secretion from -cells. In the case of norepinephrine and epinephrine, this occurs via the ₁- and ß-adrenoceptors (44, 45). There is evidence to suggest that autonomic blockade to prevent sympathetic activation results in a blunted -cell response to hypoglycemia in different species, including man (for review, see Ref. 43), although this point is still debated with a large body of evidence to the contrary. For example, the denervated dog pancreas (Ref. 46 and references therein) and the denervated human pancreas (47) secrete glucagon in response to hypoglycemia.

Norepinephrine and epinephrine inhibit secretion from ß- and -cells. There is some evidence that hyperglycemia will suppress sympathetic neuronal activity, reducing islet glucagon output (for review, see Ref. 48). The islets are also richly supplied with parasympathetic fibers originating from intrapancreatic, cholinergic ganglia (43, 49). Activation of the parasympathetic nervous system during hyperglycemia triggers insulin secretion from ß-cells, as well as somatostatin and PP release. At least four different neurotransmitters (acetylcholine, vasoactive intestinal polypeptide, pituitary adenyl cyclase-activating polypeptide, and gastrin-releasing polypeptide) of the parasympathetic system can stimulate glucagon secretion from islet -cells (for review, see Ref. 43), although their relative contribution to enhancing glucagon release in vivo remains to be elucidated.

	III. Paracrine, Autocrine, and Hormonal Regulation of Glucagon Secretion

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
Since the early 1970s, the mechanism underlying the regulation of glucagon secretion by glycemia has puzzled scientists. The debate continues whether -cells directly sense and respond to fluctuations in plasma glucose or whether the response is mediated by the autonomic nervous system and/or the paracrine/endocrine effects of secretory products from other islet cell types. Currently, a large body of research favors the latter "paracrine/endocrine" hypothesis. Although neuronal modulation of -cell activity is certainly operative, this is probably secondary to the inhibitory effects of ß-cell secretory products during hyperglycemia (50) and is unlikely to underlie the dysregulated -cell activity associated with the onset of type 1 diabetes, when ß-cell function begins to fail (51).

Clinical studies have shown that in type 1 diabetes, where normal ß-cell function is markedly impaired, glucose (administered either iv or orally) can actually stimulate glucagon secretion (52). A similar finding was reported for the alloxan-treated diabetic dog (53). In vitro, glucose can stimulate glucagon secretion from the dog pancreas when perfused in the retrograde direction (when ß-cells are down-stream of -cells) (54). Glucagon secretion from FACS-purified rat -cells is stimulated rather than inhibited by glucose (14, 15). In intact mouse islets, in which physiological glucose concentrations (like in rat islets) suppress glucagon secretion, supraphysiological glucose levels were recently reported to stimulate glucagon secretion (55). In accord with the observation that the absence of ß-cell secretory products leads to -cell hyperactivity, the glucagon secretory response to hypoglycemia is blunted in infants suffering from hyperinsulinemic hypoglycemia (56). Moreover, whereas pyruvate stimulates glucagon but not insulin secretion in intact rat islets, the -cell secretory response is lost after transduction of the ß-cells with the monocarboxylate transporter 1 (MCT-1). This is explained by the engineered pyruvate stimulation of ß-cell exocytosis leading to suppression of glucagon secretion (57). Likewise, malate and succinate stimulated glucagon secretion in -cells engineered to express the Na⁺-dependent dicarboxylate transporter-1, whereas the concomitant expression of Na⁺-dependent dicarboxylate transporter-1 in ß-cells within intact islets prevented glucagon release (58).

Thus, caution should be taken when interpreting the results of glucagon secretion experiments because -cells appear to be highly sensitive to ß-cell secretory products. This is true for clonal -cell lines that rarely display a pure phenotype (59). It is also pertinent for intact or dispersed islets, analyzed in a superfusion system, where the influence of "contaminating" ß-cell secretory products cannot be excluded. In all of these cases, high glucose continues to be associated with a reduction in -cell activity (60). Of interest, an early investigation into the effects of starvation on hormone release from the perfused rat or mouse pancreas revealed an alternative mechanism for the inhibition of glucagon release by high glucose that was independent of the usual requirement for ß-cell secretory products (61). This may reflect an adaptive mechanism involving the local neural network, -cells, or even the -cells themselves.

We use the terms "paracrine/endocrine" to encompass local interactions between different cells within the same islet. These interactions occur via the interstitium (paracrine) and/or the microcirculatory system (endocrine). It remains very difficult to discern between these two conduits when discussing intraislet signaling, although the release of ß-cell secretory products into the interstitium has been documented (62). Numerous factors modulate glucagon secretion, but in this review we will focus only on the major physiological regulators.

A. Insulin
Insulin has long been considered the most likely candidate of the plethora of ß-cell products secreted in response to high glucose to have an inhibitory paracrine effect on -cell activity (63). Particularly convincing are those studies where the action of endogenous insulin is: 1) blocked by antibody inclusion in in vitro studies in rat (14, 63, 64) and human (36), resulting in increased glucagon secretion (these data do not exclude the involvement of other paracrine factors in the regulation of glucagon release); 2) lost in islets of sulfonylurea receptor subunit 1 (SUR1)-deficient mice (SUR1 is a component of ATP-sensitive K⁺-channels), which no longer display glucose-stimulated insulin secretion or suppression of glucagon release (65, 66); or 3) diminished in alloxan diabetic minipigs, leading to a reduction of postprandial insulin-driven suppression of glucagon secretion (67). In addition, exogenously added insulin has been shown to inhibit glucagon secretion from: 1) the perfused pancreas of streptozotocin-treated rats (68); 2) alloxan-treated diabetic dogs (53, 69) and streptozotocin-treated hamsters (70); 3) islets of streptozotocin-treated guinea pigs (71); 4) isolated rat -cells (14, 60); and 5) rodent islets in low but not high glucose conditions (14, 60).

In humans, it has been reported that intraislet hyperinsulinemia prevents the glucagon response to hypoglycemia (despite an intact autonomic response) (72). It was also shown that suppression of baseline insulin secretion, abolishing the decrement in intraislet insulin during the induction of hypoglycemia, reduced the glucagon response to hypoglycemia (73). Furthermore, increasing the decrement in insulin secretion improves glucagon responses to hypoglycemia in advanced type 2 diabetes (74). These observations lead to the ß-cell "switch-off" hypothesis, suggesting that a sudden cessation of insulin secretion from the ß-cells into the portal circulation of the islet during hypoglycemia is a necessary signal for the glucagon response from downstream -cells. Support for the ß-cell switch-off hypothesis has also been obtained in streptozotocin-treated rats (75) as well as in isolated rat and human islets and islets from streptozotocin-administered rats (76). It has been argued that in several of these studies supraphysiological concentrations of exogenous insulin were required to see an effect on glucagon release. However, a physiological concentration of insulin (0.3 mU/ml) was sufficient to inhibit glucagon secretion (by about 30%) from the rat pancreas perfused in the retrograde direction in the presence of 5 mM glucose, a condition in which downstream effects of endogenous insulin are negligible (54).

Recent investigations have provided some insight into the mechanism by which insulin inhibits -cell glucagon secretion. The transcript encoding the insulin receptor is highly abundant in rat -cells, similar to another major insulin target tissue, the liver (14). Insulin receptors are also expressed in TC6 and In R1 G9 cells, and glucagon secretion was decreased with the addition of insulin in both cell types (77). Insulin transiently inhibits electrical activity and glucagon secretion in isolated rat -cells, most probably by a signaling pathway that triggers activation of ATP-sensitive K⁺-channels (K_ATP-channel) and thus membrane hyperpolarization (14). This is supported by evidence in mouse -cells where insulin activates K_ATP-channels by markedly reducing the sensitivity of the channels to ATP, an effect that is mediated via the phosphatidylinositol 3-kinase (PI3K) signaling pathway (78). Exogenously added insulin also inhibits the spontaneous Ca²⁺ oscillations observed under low glucose conditions in -cells from dissociated mouse islets (60). A similar effect was observed in TC1–9 cells and was prevented by the inclusion of the PI3K inhibitor wortmannin (60). Insulin is also known to activate K_ATP-channels in isolated mouse ß-cells (79) and in rodent hypothalamic neurons by PI3K-dependent pathways (80, 81). Insulin has been reported to activate GABA_A receptors in the -cells by receptor translocation via an AKT kinase-dependent pathway. This leads to membrane hyperpolarization in the -cell and suppression of glucagon secretion (82). The relative contribution of K_ATP-channel activation and GABA_A receptor translocation and activation to insulin-induced inhibition of glucagon release is currently unknown. Consistent with the idea that the inhibitory effect of insulin on glucagon secretion involves modulation of ion channel activity rather than granular exocytosis, epinephrine-evoked glucagon secretion from isolated rat -cells was inhibited by exogenous insulin via a cAMP-independent mechanism (83).

B. Zinc
A relative newcomer to the islet paracrine signaling scene is the metal ion Zn²⁺. Zinc cocrystallizes with insulin in ß-cell granules (84, 85, 86) and is secreted from rat (14) and mouse (87) ß-cells on exposure to high glucose (1 pmol/islet/h, a 6.5-fold increase over basal levels) (14). High local concentrations of Zn²⁺ (µM) are therefore anticipated within the islet microvasculature during hyperglycemia. The possibility that ß-cell-derived Zn²⁺ could have inhibitory effects on rat -cell glucagon release was first raised in 2003 (57) when it was found that the Zn²⁺ chelator Ca²⁺-EDTA permitted a stimulatory effect of the mitochondrial substrate monomethylsuccinate on glucagon secretion in the perfused rat pancreas, without altering the stimulated rate of insulin secretion. This was substantiated by the demonstration that exogenous Zn²⁺ reversibly inhibited pyruvate- or glucose-stimulated glucagon release from isolated rat -cells, and a mechanism for this effect has now been provided (14). Zn²⁺ can reversibly activate K_ATP-channels (EC₅₀ = 2.2 µM) in isolated rat -cells, concordantly reducing electrical activity and glucagon secretion.

Zn²⁺ has been shown to play a similar role in the central nervous system (CNS), where low concentrations can activate K_ATP-channels to restore membrane potential and inhibit glutamate release in rat neurons, playing a potentially neuroprotective role (88). Activation of K_ATP-channels in the ß-cell lines RINm5F and INS-1E by Zn²⁺ (EC₅₀ = 1.7 µM) has also been demonstrated (89, 90), and the site of action has recently been located to several histidine residues on the extracellular side of the SUR1 subunit of the K_ATP-channel (90). Interestingly, exogenously added Zn²⁺ (30 µM) has no effect on glucagon release from mouse islets in static incubations in the presence of low or high glucose (60), although Zn²⁺ (0.3–3 µM) appears to have an inhibitory effect on glucagon release from the perfused mouse pancreas in low glucose conditions (I. Franklin and C. B. Wollheim, unpublished observations). The effect of Zn²⁺ on glucagon release from human -cells remains to be investigated. The apparent convergence of the mechanisms behind both insulin and Zn²⁺ inhibition of glucagon secretion on K_ATP-channel activity is of interest and supports the idea that modulation of ion channel activity would permit paracrine signals to have a rapid and precise effect on hormone secretion (91).

C. GABA
-Aminobutyric acid (GABA) is a major nonpeptidal neurotransmitter that inhibits neuronal firing in the CNS, where it acts on two types of receptors (92). Activation of GABA_A receptors typically induces an inward Cl^– current that causes cell inhibition by hyperpolarization. Activation of GABA_B receptors usually leads to cell inhibition through closure of Ca²⁺-channels or opening of K⁺-channels (92). GABA is also produced in the ß-cells of the endocrine pancreas, where it is taken up into synaptic-like microvesicles and secreted in a regulated manner (for review, see Ref. 93). Although the circumstances under which ß-cell GABA is released only began to be elucidated recently (94, 95), it has been reported that rat islet GABA directly inhibits glucagon secretion in static assays by hyperpolarizing the -cell plasma membrane after activation of the GABA_A receptor (96). Exogenous GABA inhibits arginine-induced glucagon secretion in mouse (97) and guinea pig islets (98) as well as in isolated rat -cells (15) and TC6 cells (99). GABA_A receptors have been identified in guinea pig and rat -cells but not in rat ß- and -cells (96, 98). GABA_A receptors are heteromultimeric channels typically composed of two , two ß, and a varying third subunit. RT-PCR analysis detected transcripts of ₁ and ₄ as well as ß_1–3 GABA_A receptor subunits in rat -cells (96). It has been proposed that activation of the -cell GABA_A-receptor channel by GABA released from adjacent ß-cells in response to glucose may provide a mechanism for paracrine control of glucagon secretion (96, 98). However, when purified rat ß-cells were incubated over 24 h, high glucose concentrations inhibited GABA release into the medium, thus dissociating GABA from the stimulation of insulin secretion (100). The action of glucose is due to inhibition of mitochondrial GABA transaminase and is probably immediate (101). These results argue against a primary role for ß-cell GABA in glucose suppression of glucagon secretion in islets.

GABA_B receptor subunits have also been detected in preparations of purified rat -cells using RT-PCR (94). However, the GABA_B receptor agonist baclofen affected neither glucagon secretion from isolated rat islets nor epinephrine-stimulated exocytosis in single rat -cells (102). Further experimentation is required to determine whether protection from GABA receptor activation can alleviate inhibition of glucagon secretion during high glucose conditions in the perfused pancreas and in vivo to substantiate the physiological importance of intraislet GABA in the suppression of glucagon secretion.

D. Glutamate
L-Glutamate is the major excitatory neurotransmitter in the CNS. Endocrine cells of pancreatic islets possess glutamatergic signaling features of their own (103). -, ß-, and -Cells, to varying extents, appear to have the ability to produce, secrete, and respond to L-glutamate. L-Glutamate is produced in ß-cell mitochondria by the enzyme glutamate dehydrogenase and by cytosolic glutaminase as well as transaminase reactions in both cellular compartments. In addition to an intracellular signaling role (104), ß-cell glutamate probably accumulates in secretory vesicles and could therefore be an important intercellular signaling molecule. Vesicular glutamate transporter subtype 1 and subtype 2 have been detected in -cell secretory granules (105, 106), and indeed a parallel increase in secretion of glutamate and glucagon from intact rat islets, in response to low glucose, has been reported (106). Ca²⁺-dependent exocytosis of L-glutamate has also been observed in TC6 cells (107). Non-ß-cells have the ability to take up extracellular glutamate via a high-affinity glutamate/aspartate transport system (108, 109).

Many of the subunits that comprise the large family of glutamate receptors have been identified in the different islet cells. -Cells express both ionotropic receptors, including the AMPA and kainite subtypes (110), and metabotropic receptors (105, 111). Ionotropic glutamate receptors function as Na⁺ (but not Ca²⁺)-conducting ion channels, and their activation would therefore modulate plasma membrane potential and glucagon secretion. G protein-coupled metabotropic receptors modulate cellular cAMP levels in neurons and are typically associated with autocrine feedback inhibition of L-glutamate signaling. ß-Cells in rat and human islets also express a wide range of glutamate receptor subtypes (105, 112, 113, 114, 115, 116), and ionotropic receptor subunits have also been detected in rat -cells (117).

Given the potential complexity of intercellular glutamate signaling in the islets, what role can we anticipate glutamate playing in the regulation of -cell glucagon secretion? Activation of AMPA or kainite ionotropic glutamate receptors in low-glucose conditions stimulates glucagon release from the perfused rat pancreas (118). Conversely, at stimulatory glucose concentrations, AMPA and kainite trigger insulin secretion in vivo (119), in the perfused rat pancreas (120), and in isolated rodent islets (109, 113). In dispersed rat islets, AMPA and kainite cause membrane depolarization and Ca²⁺ influx through voltage-dependent Ca²⁺ channels (110, 113). These findings lead us to predict that glutamate cosecreted with glucagon would have a positive feedback effect on glucagon release. Similarly, we can expect glutamate released from ß-cells to promote -cell exocytosis. Although, as discussed elsewhere, ß-cell activation is synonymous with inhibition of glucagon release, any stimulatory effect of ß-cell glutamate may be masked by the inhibitory effect of other cosecreted factors, including insulin and Zn²⁺. Surprisingly, stimulation of -cell metabotropic receptors inhibited glucagon secretion from rat islets in low-glucose conditions (105), in accord with similar findings by other groups (111). The apparently opposing effects of ionotropic and metabotropic glutamate receptor activation on -cell glucagon secretion add another layer of complexity to the puzzle of intraislet glutamate signaling. Studies are now required to firmly establish the conditions under which glutamate is released from purified primary islet cells. Follow-up analyses of glutamate (and its receptor subtype-specific analogs) effects on electrical activity, cytosolic Ca²⁺ levels, and hormone secretion should be performed in isolated purified cells. Eventually, tissue-specific glutamate receptor mutagenesis studies in more intact systems such as the perfused rat pancreas may yield definitive information on the role of islet glutamate as a paracrine/autocrine effector of glucagon secretion.

E. Somatostatin
Somatostatin is a peptide hormone produced by neural and endocrine tissues, including the -cells of the endocrine pancreas. It is well established that somatostatin is a potent inhibitor of glucagon and insulin secretion in rat, dog, and human (121, 122). Somatostatin might inhibit - and ß-cell activity via the local islet microcirculatory network and/or the interstitium. These possibilities are discussed below.

Of the five distinct somatostatin receptor subtypes characterized to date, all are coupled to inhibitory G proteins and display a tissue-specific pattern of expression (123). Somatostatin receptors are expressed by islet - and ß-cells, with some conserved differences in subtype specificity. The significance of the subtype specific pattern of expression is unknown. Human, rat, and mouse -cells predominantly express the type 2 somatostatin receptor (124, 125, 126, 127, 128). In the case of human and rat ß-cells, they express predominantly type 1 (124) or type 5 (127) somatostatin receptors, respectively, in preference to type 2. However, there probably is not any absolute cell-type specificity for any particular receptor subtype, and indeed some receptor subtypes appear common to all islet endocrine cells, e.g., somatostatin receptor 5 in human islets (124).

There are at least three different mechanisms by which -cell somatostatin receptor activation leads to inhibition of glucagon secretion. Electrophysiological recordings showed that somatostatin activated G protein-coupled K⁺-channels in single rat and mouse -cells, causing hyperpolarization of the plasma membrane and inhibition of electrical activity (129, 130). The ability of somatostatin to inhibit -cell exocytosis appears dependent on the activity of a pertussis toxin-sensitive G protein (G_i2) and also calcineurin (131). Somatostatin receptor activation also inhibits adenylate cyclase activity, thereby reducing cAMP levels and protein kinase A (PKA)-stimulated glucagon secretion (132, 133, 134). Specific agonists for the type 2 somatostatin receptor were found to selectively inhibit glucagon secretion from mouse (135) and rat islets without affecting insulin release (136). Moreover, islets of transgenic mice unable to express the somatostatin receptor type 2 exhibited a 2-fold increase in arginine/K⁺-stimulated glucagon secretion, whereas basal glucagon release and glucose-induced insulin secretion did not differ from control in static assays (137). In animal models of type 2 diabetes in the nonfasted state, circulating glucagon and glucose levels were decreased after treatment with a somatostatin receptor subtype-2 agonist. In the fasted state, the agonist lowered blood glucose by approximately 25% (135). The somatostatin receptor subtype-2 agonist did not have any effects on glucagon or glucose levels in somatostatin receptor 2-deficient mice (135). This indicates that the type 2 receptor is required for the action of somatostatin on mouse glucagon secretion. These studies suggest that the inhibitory effects of somatostatin on glucagon secretion are direct and are mediated by the type 2 somatostatin receptor in rodents. However, recent data suggest that more than one somatostatin receptor subtype is likely to be involved in the regulation of glucagon secretion from rat islets (138).

Early studies in the perfused rat pancreas with antibodies against insulin, glucagon, or somatostatin led to the concept of ß directional flow of the islet microcirculation (34). The observation that perfusion with somatostatin antibody in the anterograde direction had no effect on glucagon (or insulin) secretion, but that retrograde perfusion with the antibody increased glucagon release, argued against local -cell action on -cells mediated via the circulatory system. A similar elegant study in the perfused human pancreas arrived at the same conclusion (35), although more recent work by another group showed a small but significant stimulatory effect of somatostatin antibody perfusion on glucagon secretion in low-glucose conditions only (36). However, this approach does not exclude communication via the interstitium, given that the large size of antibodies probably precludes their entry into intercellular spaces (64).

More recent work in the perfused rat pancreas showed increased arginine-stimulated glucagon secretion in the presence of the peptide antagonist DC-41–33 against the type 2 somatostatin receptor (139), which supports the concept of interstitial interaction between neighboring cells. The new wave of receptor agonists and antagonists available may prove invaluable in revealing interstitial communication between islet cells, but care should be taken to confirm that these compounds do not have nonspecific effects on hormone secretion, especially in the absence of the native receptor agonist. For instance, we have observed that the somatostatin type 2 receptor antagonist DC-41–33 (139) strongly stimulated insulin secretion from isolated rat ß-cells, in the absence of contaminating somatostatin (Fig. 2).

View larger version (20K): [in this window] [in a new window]   FIG. 2. Effect of somatostatin receptor type 2 antagonist on hormone secretion from FACS-purified rat - and ß-cells. FACS-isolated rat - or ß-cells were seeded at a density of 20,000 cells per well in polyornithine-coated 24-well plates and cultured overnight. Static assays were performed to determine -cell glucagon secretion (A) or ß-cell insulin release (B) during a 30-min incubation in basal (2.5 mM glucose) or stimulatory (A, 5 mM pyruvate; B, 16 mM glucose) conditions. An antagonist against the type 2 somatostatin receptor (SstR antagonist, 2 µM) (139 ) or an antibody against somatostatin (Sst antibody) was included where indicated. Endogenous somatostatin (10 pM) was detected in the -cell but not ß-cell fraction. The addition of exogenous somatostatin (10 nM) inhibited glucose-stimulated insulin release, confirming that the ß-cell fraction was essentially free of contaminating -cells, as expected (432 ). Data are the mean ± SEM of three different experiments. *, P 0.05.

F. Ghrelin
Ghrelin, a recently described GH-releasing peptide, is produced primarily in rat or human stomach during fasting. Ghrelin also appears to be synthesized in human and rat islets (21, 154). Although, in the case of the rat, there is controversy over the presence of ghrelin-positive cells in the adult islet (22, 23) contrasting with (154, 155, 156), it seems that they are almost certainly present earlier in development, comprising a small subpopulation of cells located at the islet periphery. Whether some of these cells in the rat islet are in fact also glucagon positive remains to be clarified (compare Refs. 154, 156 , and 157 with21 and 22), although in human islets ghrelin-positive cells probably comprise a distinct peripheral subpopulation (21). Such cells have also been detected in the mouse islet (158).

The receptor to which ghrelin binds, GHS-R, is widely expressed in the adult rat islet (22, 154) and has been colocalized with glucagon-positive cells (22). Locally released ghrelin would be anticipated to act in an autocrine/paracrine fashion on -cell glucagon release, intuitively perhaps promoting glucagon release during hypoglycemia. However, studies in perfused rat pancreas indicate that concentrations of ghrelin within the physiological range (0.5–3 nM) had no effect on basal or arginine-stimulated glucagon release, despite mild inhibition of both arginine-induced somatostatin release and glucose-induced insulin secretion (159). Perhaps, under the conditions tested, locally released ghrelin was already exerting a maximal effect on -cell activity, masking any further stimulatory effect of exogenously added ghrelin. Contrasting findings have been reported in mouse, where exogenously added ghrelin increased glucagon release from islets in static assays at all glucose concentrations tested (160), although under such assay conditions indirect effects cannot be excluded. In the same study, however, ghrelin (10 nM per kg body weight) did not raise basal plasma glucagon and indeed impaired the glucagon secretory response to carbachol, 3-isobutyl-1-methylxanthine, and arginine, suggesting an inhibitory effect on secretion. GHS-R expression has not yet been reported in mouse islets. The physiological relevance of islet ghrelin needs more investigation.

Interestingly, another peptide encoded by the ghrelin gene has recently been identified. The peptide is named obestatin, and contrary to the appetite-stimulating effects of ghrelin, treatment of rats with obestatin suppressed food intake, inhibited jejunal contraction, and decreased body weight gain (161). These effects of obestatin are mediated via the orphan G protein-coupled receptor GPR39 (161). It will be interesting to investigate whether obestatin is produced and secreted from ghrelin-producing islet cells. The GPR39 receptor is expressed in pancreas (162), and future research should be directed to understanding the potential role of obestatin as well as GPR39 in islet function.

G. GLP-1
Glucagon-like-peptide (GLP)-1 is an incretin released from the L cells in the small intestine after meal ingestion. One of the primary actions of GLP-1 is the augmentation of glucose-induced insulin secretion, directly by activation of GLP-1 receptors expressed on ß-cells and indirectly via the autonomic nervous system (for review, see Ref. 163). GLP-1 is currently under intensive investigation for its potential application as an antidiabetogenic peptide. Evidence suggests that GLP-1 may also have suppressive effects on glucagon release from -cells (for review, see Ref. 164). Any inhibitory effect of GLP-1 on glucagon release may involve direct (via GLP-1 receptor expression in -cells) and/or indirect mechanisms. An indirect action could be mediated by the stimulatory effect of GLP-1 on neighboring ß- or -cells, causing intraislet paracrine inhibition of glucagon release.

Clinical studies have demonstrated that patients with type 1 diabetes have decreased plasma glucagon levels in response to iv application of GLP-1 after an overnight fast (165). However, a small increase in plasma C-peptide levels was also detected, so an indirect effect of GLP-1 via activation of residual ß-cell activity cannot be ruled out. In a different study, patients with type 1 diabetes infused acutely with insulin or insulin plus GLP-1, under euglycemic clamped conditions (5.3 mM glucose), did not demonstrate increased inhibition of plasma glucagon levels in the presence of GLP-1 (166). The possibility that the inhibitory effect of insulin alone was already maximal cannot, however, be excluded. Long-term (5 d) administration of GLP-1 to type 1 diabetic patients caused a small reduction in postprandial plasma glucagon levels (167). Studies in perfused rat pancreas (168) have shown that GLP-1 alone can inhibit glucagon secretion normally associated with a decrease in glucose concentration (11 to 3.2 mM glucose), without effecting insulin or somatostatin release, and thereby largely excluding the possibility of indirect GLP-1 action via a paracrine mechanism.

GLP-1 receptor expression in human -cells has not been investigated. Conflicting reports have been published for the rat. Neither GLP-1 receptor nor its transcript could be detected in purified rat -cells (14, 169). Direct GLP-1 application to these cells did not alter glucagon secretion (14) or cause an increase in cAMP levels (in contrast to glucose-dependent inhibitory peptide, another glucoincretin) (169). However, GLP-1 receptor expression was detected by immunocytochemistry in a subpopulation (20%) of glucagon-positive cells in dispersed rat islets (170), and GLP-1 caused an increase in the rate of exocytosis in single rat -cells (171, 172). GLP-1 was also recently reported to elicit an increase in the cAMP content of isolated rat -cells (172). On the other hand, spontaneous Ca²⁺ oscillations detected in mouse -cells of dispersed islets were abolished by the application of GLP-1 (173).

Evidence for a primarily paracrine mechanism in GLP-1 inhibition of glucagon release in the mouse has recently come to light. In static islet assays, GLP-1 had a strong suppressive effect on glucagon release (65), and plasma glucagon levels were inhibited by administration of GLP-1 (174). In both cases, a loss of ß-cell responsiveness to GLP-1, either by global knockout of the K_ATP-channel subunit SUR1 (65) or ß-cell specific knockout of the transcription factor Pdx-1 (174), resulted in a loss of -cell responsiveness to GLP-1. It should be noted that, in the former case, the requirement for functional -cell K_ATP-channels for a direct inhibitory effect of GLP-1 on glucagon release cannot be ruled out (65).

The third, and as yet unexplored possibility, is that GLP-1 inhibition of glucagon secretion is mediated by local neural networks. GLP-1 is rapidly degraded in the circulatory system. Indeed, reports indicate that probably a substantial component of the potentiating effect of GLP-1 on glucose-induced insulin secretion is mediated by the autonomic nervous system in rats (175) and mice (163, 176), possibly via GLP-1 activation of the vagal hepatic nerves (177, 178). Neuronal control of glucagon secretion is well documented (see below) and could provide an explanation for the disparity between in vivo or pancreatic perfusion data (local neural networks intact) and the apparent absence (or at least very weak expression) of the GLP-1 receptor on -cells. We refer the reader to a recent review by Dunning et al. (164) for effects of GLP-1 on -cell function and glucagon release in healthy subjects and in diabetic patients.

H. Glucagon
Glucagon stimulates insulin and somatostatin secretion from perfused rat and human pancreata (36, 148). The autocrine effect of secreted glucagon on rat and mouse -cell activity has only recently been examined (172). The authors demonstrated glucagon receptor expression and a stimulatory effect of glucagon on cAMP levels and exocytosis in isolated mouse and rat -cells, suggesting a positive feedback effect. Transgenic mice unable either to produce glucagon (pro-convertase 2 knock-out) (179) or to express the glucagon receptor (180) exhibit -cell hyperplasia, indicating that either the loss of autocrine action or perhaps the absence of glucagon action in target tissues leads to -cell hyperplasia. Interestingly, in antisense oligonucleotide studies where glucagon receptor expression was specifically reduced in mouse liver, an increase in glucagon release from -cells in vivo was detected in the absence of an increase in -cell mass (181). The signal that triggers -cell hyperplasia may therefore indeed be autocrine, rather than the result of reduced glucagon action on the liver per se. PP cells substitute for glucagon-positive cells in the islets of the ventral rat pancreas (182). However, no effect of exogenous PP on glucagon secretion from the perfused rat pancreas was observed in low- or high-glucose conditions, despite an inhibitory effect on insulin secretion (83, 183).

	IV. Autonomic Regulation of Glucagon Secretion

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
Increased glucagon release represents the most important glucose counterregulatory factor that is critical to prevent or rapidly correct hypoglycemia. Glucose is an essential metabolic fuel for the brain under physiological conditions. Because the brain cannot synthesize glucose or store more than a few minutes’ worth of glucose consumption as glycogen, it is critically dependent on a continuous supply of glucose from the circulating blood (184). Falling arterial glucose concentrations are sensed in the hepatic portal vein, the carotid body, and different areas of the brain. One brain region in particular, the VMH, has been shown to play an important role not only in sensing decreases in blood glucose levels but also in initiating counterregulatory responses to hypoglycemia (16, 185, 186, 187, 188). As arterial plasma glucose levels decline in the physiological range, insulin secretion decreases mainly due to the importance of glucokinase-mediated glucose sensing in the ß-cell. This will enhance hepatic glucose production. A fall in plasma glucose concentration below the physiological range stimulates the secretion of glucagon and epinephrine. Glucagon stimulates hepatic glycogenolysis as well as gluconeogenesis. Epinephrine enhances hepatic glucose output and decreases glucose clearance by tissues such as muscle and fat. Epinephrine also mobilizes gluconeogenic precursors including lactate, amino acids, and glycerol (see Fig. 1).

Two main mechanisms have been proposed to mediate the -cell response to hypoglycemia (Fig. 3). The first mechanism involves relief from inhibitory paracrine/endocrine influences by neighboring ß- and -cells as discussed above. The second mediator of the glucagon release to hypoglycemia is the autonomic input to the -cell (Fig. 3). There are three major autonomic influences on the -cell: sympathetic nerves, parasympathetic nerves, and circulating epinephrine. All three autonomic inputs are activated by hypoglycemia and are capable of stimulating glucagon secretion (for review, see Ref. 43). However, the relative contribution of the different autonomic inputs to enhance glucagon secretion during hypoglycemia is unknown. This is emphasized by the observations that the denervated dog (46) and human (47) pancreas release glucagon in response to hypoglycemia. The evidence for autonomic mediation of glucagon secretion during hypoglycemia has been reviewed extensively (43, 184, 189).

View larger version (24K): [in this window] [in a new window]   FIG. 3. Mechanisms for regulation of glucagon release in response to hypoglycemia. A, During euglycemia, glucagon secretion may be suppressed (thin arrow) as a result of: 1) the lack of autonomic stimulation, including adrenergic stimulation of the -cell; and 2) a marked paracrine inhibition by factors released from ß-cells (and -cells, not indicated in figure) within the islet. Green arrows indicate stimulatory pathways, whereas red arrows symbolize inhibitory pathways. B, During hypoglycemia, glucagon secretion is markedly increased (thick arrow). This may arise from a marked reduction of the paracrine inhibitory effects (including 2-adrenergic receptor activation in ß-cells) as well as a stimulatory action of the autonomic nervous system (involving ß-adrenergic receptor activation in the -cell) secondary to its activation by central hypoglycemia.

The involvement of K_ATP-channels in the glucagon counterregulatory response is supported by the observation that VMH neurons in Kir6.2-deficient mice, which lack functional neuroendocrine-type K_ATP-channels, have a persistently elevated firing rate and exhibit a severely impaired glucagon counterregulatory response (188). However, it is important to emphasize that the mouse is a whole-body Kir6.2 knockout and therefore the contribution of VMH neurons to the impaired glucagon counterregulatory response is unknown. Pharmacological activation of K_ATP-channels in the VMH amplified the counterregulatory glucagon and epinephrine responses to hypoglycemia (194), whereas K_ATP-channel closure by sulfonylureas suppressed the counterregulatory responses to hypoglycemia (16). This is consistent with increased activity of the autonomic nervous system and stimulation of glucagon secretion in glucose transporter 2-deficient mice in the fed state and suppressed glucagon secretory response to hypoglycemia (195, 196). This dysregulation was caused by inactivation of centrally located glucose sensors that require expression of glucose transporter 2 in glial cells and control the activation of the nucleus of the tractus solitarius and dorsal motor nucleus of the vagus nerve (197). Neurons in the nucleus of the tractus solitarius are sensitive to small variations in blood glucose concentrations (198) and send projections to the hypothalamic nuclei as well as the dorsal motor nucleus of the vagus and are connected to the pancreas (199). Gene deletion of the glial cell line-derived factor family receptor 2 causes islet parasympathetic denervation and results in marked blunting of the glucagon secretory response to neuroglucopenic stimulation after injection of 2-deoxy-glucose (49). In man, section of the abdominal vagus trunc (vagal truncotomy) blunts the glucagon secretory response to induced hypoglycemia. Surprisingly, selective vagotomy (section of the branches to the pancreas) did not alter the glucagon response (200). This suggests that the response is mediated at least in part by stimulation of the splanchnic sympathetic nerves. Glucose recognition in the CNS has been suggested to involve metabolic coupling between astrocytes and neurons (201, 202). This model suggests that glucose is initially taken up by astrocytes and metabolized to lactate, which is then transported into neurons for ATP generation and the control of nerve firing rate.

As outlined above, several paracrine/endocrine factors as well as glucose-sensing neurons in the VMH have the potential to modulate glucagon secretion. However, the physiological importance of some of these factors remain to be established, as well as whether they exert their actions via direct effects on the -cell or by intraislet paracrine/endocrine mechanisms. The main modulators of -cell secretion and their proposed signaling mechanisms are outlined in Fig. 4A. Binding of insulin to its receptor causes activation of K_ATP-channels resulting in membrane hyperpolarization and inhibition of glucagon secretion. Insulin may also cause translocation to and activation of GABA_A receptors in the -cell plasma membrane. Additionally, K_ATP-channels are activated by Zn²⁺, which is coreleased with insulin from the ß-cell. Although it is well established that GABA_A and somatostatin receptor activation results in -cell membrane hyperpolarization, the physiological significance of intraislet GABA and somatostatin for regulation of glucagon release remains to be fully established. Circulating epinephrine stimulates adenylate cyclase activity, an increase in intracellular Ca²⁺ concentration ([Ca²⁺]_i), and glucagon release. Finally, activation of intraislet parasympathetic nerve endings stimulates glucagon secretion by release of a cocktail of different neurotransmitters (Fig. 4A). The effects of the main modulators on glucagon secretion from FACS-isolated rat -cells are depicted in Fig. 4B. The effects of glucose, pyruvate, arginine, tolbutamide, and diazoxide on glucagon release from rat -cells (Fig. 4B) are also shown for completeness.

View larger version (34K): [in this window] [in a new window]   FIG. 4. A schematic overview of main physiological regulators of -cell function and glucagon secretion. A, Model summarizing site of action and intracellular signaling mechanisms for main physiological modulators of -cell stimulus-secretion coupling. Green dotted lines depict activation, whereas red dotted lines represent inhibition of ion channel activity or intracellular signaling pathways. AC, Adenylate cyclase; Gs, stimulatory G protein; Gi, inhibitory G protein; PNS, parasympathetic nervous system. B, Glucagon secretion from FACS-isolated rat -cells exposed for 30 min to Krebs-Ringer bicarbonate HEPES buffer (15 ) containing either 1 mM or 20 mM glucose. Where indicated, the extracellular medium was supplemented with pyruvate (5 mM), arginine (10 mM), Zn2+ (30 µM), or epinephrine (5 µM). Insulin, GABA, and somatostatin were tested at a concentration of 100 nM, whereas tolbutamide and diazoxide were applied at 100 µM. Data are mean ± SE of five to nine different experiments. *, P < 0.05; **, P < 0.01.

	V. Transcriptional Control of Pancreatic -Cell Development

Top Abstract I. Introduction II. Islet Endocrine Cell... III. Paracrine, Autocrine, and... IV. Autonomic Regulation of... V. Transcriptional Control of... VI. The Glucagon Gene:... VII. {alpha}-Cell Stimulus... VIII. {alpha}-Cell... IX. Summary and Conclusions Note Added in Proof References

 
During embryogenesis, the pancreas develops by fusion of a dorsal and ventral epithelial bud generated by evagination of the foregut (for review, see Ref. 203). The induction of the pancreatic phenotype in the dorsal pancreatic bud has been suggested to be dependent on signals from the adjacent notochord (204, 205), although more recent evidence suggests that this signal originates from blood vessels (206, 207). In the mouse, the dorsal bud appears on embryonic day 9.5 (E9.5) concomitantly with the first differentiated glucagon-producing -cells (208, 209, 210). The following day (E10.5), insulin-producing cells are detected, often coexpressing glucagon (209). At E14, numerous fully differentiated ß-cells appear and the endocrine cells start to become organized in small aggregates, and within the next 24 h, the first somatostatin-producing -cells become visible. Finally, shortly before birth, PP-producing cells differentiate, and the endocrine cells begin to form well-organized islets of Langerhans.

Several transcription factors have been identified based on their temporally and spatially restricted expression during pancreatic development. Furthermore, lineage studies and generation of mice with targeted mutations of the genes that encode these factors have enhanced our understanding of islet and -cell differentiation (Table 1). Deletion of pTF1a/p48, a basic helix-loop-helix transcription factor expressed early in endoderm and later in the mature exocrine pancreas, results in pancreagenesis (211). Similarly, gene inactivation of Pdx-1, a homeobox gene expressed in pancreatic buds, or of the LIM homeodomain gene Isl-1 also results in complete absence of pancreas development (212, 213, 214). Mice lacking Neurogenin 3 (Ngn3), a basic helix-loop-helix (bHLH) transcription factor, fail to develop endocrine cells (215). The proendocrine function for Ngn3 is supported by lineage studies (216) and ectopic expression of Ngn3 under the control of the Pdx-1 promoter, which leads to premature differentiation of the pancreatic progenitor cells into endocrine cells, although predominantly glucagon-producing cells (217, 218). Thus, whereas Ngn3 activation in pancreatic progenitor cells promotes an endocrine commitment, the specification of the different islet cell types is controlled by other transcription factors (for review, see Refs. 219 and 220) (Fig. 5). Pax-6 is specifically involved in the differentiation of -cells because no or few glucagon-producing cells are observed in homozygous mutants (221). Brain-4 (Brn-4), a POU-homeodomain-containing protein, is expressed in the pancreatic anlage of the mouse foregut at E10 in glucagon-producing cells and seems to transactivate glucagon gene expression (222). However, loss-of-function mutant mice do not exhibit any defect in -cell formation or function (223), although recent in vivo and in vitro studies have shown that forced expression of Brn-4 is sufficient to activate the glucagon gene but not a complete -cell phenotype (224, 225). This suggests a role for Brn-4 in the islet to promote hormone expression (Fig. 5). The cell type-restricted bHLH transcription factor Beta2/Neu