Glucagon and the Glucagon-Like Peptides

Published on 28/03/2015 by admin

Filed under Endocrinology, Diabetes and Metabolism

Last modified 22/04/2025

Print this page

This article have been viewed 3315 times

Chapter 9

Glucagon and the Glucagon-Like Peptides

Daniel J. Drucker

Glucagon and the glucagon-like peptides (GLPs) are derived from a single proglucagon gene in mammals and exhibit an increasing number of biologically important actions. Glucagon, synthesized principally in islet α cells, is a key regulator of glucose homeostasis through its actions on enzymes that regulate glucose production and glycogen synthesis. GLP-1 and GLP-2 are liberated from intestinal endocrine cells and regulate glucose homeostasis and intestinal epithelial growth, respectively. These three peptides exert their actions via interaction with unique receptors that exhibit distinct patterns of tissue-specific expression. This chapter reviews our current understanding of the biology of glucagon and the proglucagon-derived peptides (PGDPs), emphasizing recent advances in our understanding of glucagon biosynthesis and action and the therapeutic potential of the glucagon-like peptides.

Biosynthesis of Pancreatic Glucagon

Proglucagon is encoded by a single gene in mammals that gives rise to a proglucagon mRNA with major sites of expression in pancreatic islets, brain, and enteroendocrine cells of the small and large intestine. Proglucagon also may be expressed in subsets of salivary glands and taste buds.¹ The proglucagon gene contains six exons, several of which encode distinct functional peptide domains (Fig. 9-1). Glucagon, a 29 amino acid peptide, is synthesized in the A cells of the pancreatic islets of Langerhans. Islet A cells are distinguishable from insulin-producing β cells in part by the morphology of their respective secretory granules and by their anatomic distribution, predominantly at the periphery of the islet. The peripheral location of islet A cells, together with functional studies demonstrating central (from a core of β cells) to peripheral (A cells) islet blood flow, raises the possibility of a tightly regulated islet microenvironment. Nevertheless, as the distribution of islet α and β cells may vary from species to species, the functional importance of islet cell distribution remains unclear.

FIGURE 9-1 Structure of proglucagon and the proglucagon-derived peptides. The amino acid sequences of glucagon, glucagon-like peptide (GLP)-1, and GLP-2 are shown, as well as the cleavage site for dipeptidyl peptidase-4 (DPP-4). IP-1 and IP-2, Intervening peptides-1 and -2, respectively.

Islet Transcription Factors and the α Cell

The pancreas derives from the upper foregut, with dorsal and ventral pancreatic buds giving rise to pancreatic epithelium that eventually forms the exocrine and endocrine pancreas. This process involves signals from mesenchyme and requires the correct temporal expression of growth factors, transcription factors, and related signaling molecules. Pancreatic islet cells appear to be of endodermal origin and are first detectable by embryonic day 8.5 (E8.5) in the mouse. The first islet hormone-immunopositive cells detectable in developing islets contain glucagon immunoreactivity. Insulin immunopositive cells are detectable by day E14 in the mouse, followed several days later by the appearance of somatostatin-containing cell types.

Studies of targeted gene disruption in mice have provided new insights into the organization of islet endocrine cells in the pancreas. Neurogenin-3 is critical for pancreatic islet cell development and differentiation and controls the expression of many genes important for α cell differentiation and glucagon gene transcription. Ngn3 interacts in a regulatory network with the zinc-finger protein Myt1 to promote glucagon expression in α cells.² Mice with a null allele in the basic helix-loop-helix transcription factor p48 fail to develop exocrine pancreatic tissue; however, hormone-secreting islet cells, including glucagon-producing α cells, are found within the mesentery during embryonic development, and later in the spleen.³ A key role for cell adhesion molecules in the control of spatial organization of islet cells is illustrated by analysis of the endocrine pancreas in mice expressing a dominant negative E-cadherin receptor in islet β cells. These mice exhibit abnormal clustering of β cells, yet glucagon-producing α cells are still capable of aggregating into islet-like clusters.⁴ In contrast, the normal peripheral distribution of islet α cells is markedly perturbed in neural cell adhesion molecule (NCAM)-/- mice; however, the number of α cells and the glucagon content of the pancreas remain unaffected.⁵ It is intriguing that mice with genetic disruption of the ciliogenic transcription factor regulatory factor X (RFX)-3 exhibit small disorganized islets with markedly reduced insulin and glucagon content, implying a role for cilia in optimal islet organization and cell development.⁶

Considerable insight into the developmental biology of the endocrine pancreas and islet α cells has been derived from studies of islet transcription factors in cell lines and mice. Gene knockout studies demonstrate that the LIM domain protein isl-1 is necessary for the formation of differentiated islet cells, including α cells, in the developing pancreas.⁷ Targeted deletion of the homeobox transcription factor Arx results in mice that exhibit complete failure of α cell development, leading to glucagon deficiency and neonatal hypoglycemia.⁸ In contrast, although mice deficient in the homeobox transcription factor Ipf/Pdx-1 fail to develop a pancreas, a few islet cells immunopositive for insulin or glucagon are observed in Ipf/Pdx-1-/- mice at E11.⁹ These observations suggest that Ipf/Pdx-1 is not essential for the formation of islet α cells, whereas loss of Pdx-1 expression appears to be required for acquisition of a differentiated α cell phenotype.

The homeobox transcription factor Nkx2.2 is expressed in adult α, β, and PP islet cells; mice with targeted disruption of NKx2.2 lack β cells, develop diabetes, and exhibit a marked reduction in the numbers of islet α cells.¹⁰ A related phenotype is observed in NeuroD/BETA2-/- mice, which exhibit marked reductions in the numbers of islet β and α cells and develop diabetes shortly after birth,¹¹ a phenotype dependent on the specific genetic background. Similarly, disruption of pax6 function in mice results in poorly formed islets with disorganized islet architecture and markedly reduced but detectable numbers of both β and islet α cells.^12,13 It is intriguing that mice that harbor mutations in islet transcription factors also may exhibit paradoxically increased numbers of islet α cells. For example, targeted deletion of the Pax4 gene results in poorly formed islets, with a marked reduction in islet β cells and comparatively increased numbers of islet α cells.¹⁴ A related phenotype is observed in mice with a homozygous deletion in the glucose transporter 2 (GLUT-2) gene, with a marked increase in the ratio of α to β cells observed in GLUT2-/-islets.¹⁵ Whether these findings are due to a block in the normal islet differentiation program, or to loss of inhibitors that restrain α cell proliferation, is not known. Evidence in support of glucagon itself regulating the numbers of islet α cells derives from analysis of mice with targeted disruption of the genes that encode prohormone convertase-2, or the glucagon receptor. These mice represent models for loss of bioactive glucagon¹⁶ or glucagon action,¹⁷ respectively, and they exhibit marked α cell hyperplasia, which is corrected in the former instance by glucagon replacement therapy.

Proglucagon Gene Transcription Factors

A combination of gene transfection experiments using cell lines in vitro and transgenic studies of promoter function in vivo has yielded insight into the molecular control of glucagon gene transcription. Differential gene expression analysis has revealed a remarkably complex pattern of transcription factors preferentially expressed in α versus β cell lines ¹⁸ and in single α cells isolated from the developing endocrine pancreas.¹⁹ The rat proglucagon gene promoter directs transcriptional initiation from a single transcription start site that maps to an identical location in brain, pancreas, and intestine.²⁰ Cell transfection experiments utilizing islet cell lines have identified five distinct regions, designated G1 through G5, within the proximal proglucagon promoter that exhibit functional importance for activation of islet cell–specific proglucagon gene transcription. The proximal G1 region is AT rich, interacts with both widely expressed and islet cell–specific proteins, and is functionally important for specifying expression in islet A cells. G1 interacts with the homeobox transcripton factors Isl-1, Cdx-2/3, Brn4, and Pax6.^13,21–24 These transcription factors bind G1 and activate reporter genes containing G1-derived sequences in transfection assays. Pax6 and c-Maf exhibit cooperativity in their ability to increase G1-dependent transcriptional activation,²⁵ whereas Nkx6.1 inhibits Pax6-dependent G1 activation. Reduction of isl-1 expression leads to reduced G1-dependent proglucagon promoter activity and decreased levels of proglucagon mRNA transcripts.²¹ Similarly, increased expression of Cdx-3 in islet cells is associated with induction of both transfected G1-dependent reporter genes and endogenous proglucagon gene expression.^22,26 The G4 element, located just upstream of G1, binds a complex resembling insulin-enhancer factor 1 (IEF1), and recent experiments suggest that IEF1 represents a heterodimer of the helix-loop-helix (HLH) proteins E47 and BETA2.²⁷

The more distal G3 region functions as an islet enhancer–like element and has been divided into two distinct functional domains. Subdomain A contains a sequence element similar to sequences found within the insulin and somatostatin promoters, leading to the designation of this composite sequence as a pancreatic islet cell–specific enhancer sequence, or PISCES element.28,29 Sequences within domain A of G3 also appear to mediate the inhibition of glucagon gene transcription by insulin.^30–32 The pax6 protein has been identified as a positive activator of proglucagon gene transcription that exerts its function in part through interaction with domain A of the G3 element.¹³ Pax6 also binds the G1 element, either as a monomer or via heterodimer formation with cdx-2/3.³³ Disruption of Pax6 expression in gene-targeted mice results in loss of islet A cells.^12,13 The levels of pancreatic and intestinal proglucagon mRNA transcripts are also markedly reduced in SEY^NEU mice that harbor a dominant negative mutation in the pax6 gene. Thus Pax6, a PISCES binding protein, is functionally important for both islet and enteroendocrine cell development and activation of proglucagon gene transcription via the G3 enhancer and G1 promoter elements.^12,13,33,34

The G2 element mediates the positive and negative actions of HNF-3 (Foxa) proteins, with isoforms of HNF-3β and HNF-3α competing for binding to the G2 element and serving as repressors and activators, respectively, of proglucagon gene transcription.35,36 HNF-3γ also binds to and transactivates the proglucagon promoter G2 element but does not appear to be essential for islet cell formation or glucagon gene transcription.^37–39 Whether HNF-3β plays an essential role in the development of islet α cell formation or glucagon gene transcription in vivo remains unknown, as targeted inactivation of the HNF-3β gene results in embryonic lethality before formation of the endocrine pancreas.^40,41 Although the numbers of α cells and the morphology of islets appear histologically normal in HNF3α-/- mice, the development of neonatal hypoglycemia in the face of inappropriately reduced levels of circulating glucagon and proglucagon mRNA transcripts demonstrates the essential role of HNF-3α in pancreatic proglucagon gene transcription.³⁷

Activation of the cyclic adenosine monophosphate (cAMP)-dependent pathway leads to transcriptional induction of proglucagon gene expression in both islet and intestinal cells through a cAMP response element (CRE) present upstream of the G3 element in the proximal promoter region of the rat proglucagon gene.42,43 The CRE also mediates transcriptional activation by ATF3 family members⁴⁴ and is the target for pharmacologic depolarization–dependent induction of rat proglucagon gene transcription.⁴⁵ A second calcium response element has been localized to the G2 element, and calcium responsiveness may be mediated via interaction of nuclear factor of activated T cells (NFAT)-like proteins with members of the HNF-3 family.⁴⁶ The sequence of the CRE element is less well conserved in the human proglucagon promoter, and its functional relevance for proglucagon gene expression in human islets has not yet been elucidated.

Transgenic experiments in mice have identified distinct DNA sequences essential for tissue-specific proglucagon gene transcription in vivo. The first 1253 nucleotides of the rat proglucagon promoter direct heterologous transgene expression to the islets and brain but not the intestine in transgenic mice.⁴⁷ In contrast, targeting of transgene expression to enteroendocrine cells in vivo requires the presence of additional rat proglucagon gene 5′-flanking sequences between −1253 and −2252.⁴⁸ Foxo1 binds to an upstream site at −1798 and appears to be important for both basal and insulin-regulated proglucagon gene transcription in α cells.⁴⁹ DNA sequences constituting the human proglucagon gene promoter exhibit different functional properties than homologous rat sequences, as the first 1600 bp of the human proglucagon gene 5′-flanking region target transgene expression to appropriate cell types in the brain and intestine, but not the pancreatic islets in transgenic mice.⁵⁰

Islet Proglucagon Biosynthesis

Consistent with the increased circulating glucagon levels in patients with diabetes, experimental diabetes in rodents is associated with increased levels of pancreatic proglucagon mRNA; correction of the insulin deficiency, but not the hyperglycemia, normalizes the levels of proglucagon mRNA.⁵¹ Paradoxically, insulin may increase proglucagon gene expression in enteroendocrine cells.⁵² Orexin-A also inhibits glucagon secretion and proglucagon gene expression via the phosphatidylinositol-3-kinase–dependent pathway, requiring Foxo1.⁵³ In contrast, few hormones or metabolites have been identified that increase pancreatic proglucagon biosynthesis in vivo. Although experiments using in situ hybridization suggested that 4 days of fasting in the rat leads to a doubling of pancreatic proglucagon mRNA transcripts,⁵⁴ hypoglycemia induced by insulin failed to demonstrate upregulation of proglucagon mRNA.^51,55 Available data suggest that regulation of glucagon secretion, and not biosynthesis, may be a more relevant locus of control in vivo.

Following transcription of proglucagon mRNA and translation of the proglucagon precursor, 29 amino acid mature glucagon is liberated via posttranslational processing by specific prohormone convertases differentially expressed in the islet α cell. In contrast to processing in the enteroendocrine cell (see Fig. 9-1), the amino acid sequences carboxyterminal to glucagon remain unprocessed and are secreted as part of a larger polypeptide designated the major proglucagon fragment (MPGF). PC2 is essential for processing of proglucagon to glucagon in α cells as PC2-/- mice exhibit hypoglycemia, α cell hyperplasia, and glucagon deficiency with an accumulation of incompletely processed PGDPs in the pancreas.¹⁶ Whether PC2 directly liberates glucagon from proglucagon, or cleaves proglucagon to glicentin, which is subsequently processed to glucagon, remains unclear. The enzyme PC1 is responsible for cleavage of proglucagon to yield an intestinal profile of PGDPs,^56,57 and PC1 null mice exhibit impaired processing of proglucagon to the glucagon-like peptides.⁵⁸ Moreover, human subjects with an inactivating mutation in the PC1 gene exhibit incompletely processed proglucagon in gut endocrine cells and deficiency of both GLP-1 and GLP-2.⁵⁹ The convertases important for proglucagon processing in the brain have not been identified definitively.

Glucagon Secretion

The secretion of glucagon is regulated, both positively and negatively, by neuropeptides, hormones, metabolites, and the autonomic nervous system. The islet α cell plays a central role in the defense of blood glucose, with hypoglycemia stimulating and hyperglycemia suppressing glucagon secretion in vivo. A cells express voltage-dependent Na⁺, K⁺, and Ca²⁺ channels that interact in the regulation of membrane potential and, ultimately, depolarization. Amino acids such as glutamine, alanine, pyruvate, and arginine stimulate both insulin and glucagon secretion, and this observation provides the basis for the use of arginine in the assessment of islet and α cell function in rodent and human physiology.⁶⁰ Both α- and β-adrenergic receptors modulate glucagon secretion. Epinephrine stimulates glucagon secretion via protein kinase A (PKA)-dependent enhancement of Ca²⁺ influx through L-type calcium channels, leading to granule exocytosis. Peptidergic activators of glucagon secretion include cholecystokinin (CCK), pituitary adenylate cyclase–activating polypeptide (PACAP), gastrin, urocortin III, vasopressin, and glucose-dependent insulinotropic peptide (GIP).

In contrast, inhibitors of glucagon secretion include glucose, somatostatin-14, and γ-aminobutyric acid (GABA).⁶¹ Somatostatin tonically inhibits glucagon release, and mice with disruption of the somatostatin receptor 2 (SSTR2) gene exhibit mild hyperglycemia and elevated levels of nonfasting glucagon.⁶² Glucose induces GABA release from β cells, providing an indirect mechanism for glucose-mediated suppression of glucagon secretion.^63–65 Glucose also reduces electrical activity and exocytosis via depolarization-induced inactivation of ion channels, specifically KATP, Na⁽⁺⁾ (TTX) and N-type Ca⁽²⁺⁾ channels, as a direct mechanism for inhibition of glucagon secretion. Insulin also may directly inhibit glucagon release, possibly via insulin receptors expressed on α cells. The anatomic arrangement of peripheral islet α cells surrounding a core of β cells, together with the functional results of immunoneutralization studies, provides additional evidence for intra-islet insulin inhibiting downstream of α cells in some species.⁶⁶ Although the mechanisms underlying the paracrine inhibition of α cell activity remain unclear, experimental evidence supports an important role for β cell secretory products such as zinc as negative regulators of glucagon secretion.^65,67

The effects of glucose on glucagon secretion are integrated with the inhibitory effects of insulin on α cell secretion, as hyperglycemia stimulates insulin secretion, whereas a drop in blood glucose suppresses insulin release from the β cell, thereby relieving the α cell from the tonic inhibitory actions of insulin. These actions are exemplified by meal ingestion, which is associated with increased circulating nutrients and insulin secretion, and reduced levels of circulating glucagon. In contrast, blood glucose is maintained in the fasting state by hepatic glucose production due in part to increased levels of glucagon secretion and suppression of insulin release from the β cell.68,69 Consistent with the effects of individual nutrients on α cell function, infusion of arginine stimulates glucagon secretion in both normal subjects and individuals with diabetes mellitus.⁶⁹ The response to arginine infusion in normal human subjects, namely, stimulation of insulin and glucagon release, is reproducible, and as expected, the insulin response is significantly greater and the glucagon response attenuated as the glucose concentration increases.⁷⁰

In normal subjects, glucagon secretion rises as glucose falls in the fasting state and may be further stimulated by exercise, consistent with the physiologic role for glucagon in regulating glucose production. The levels of both glucagon and plasma catecholamines increase with graded exercise in normal human subjects, with the greatest increments in plasma glucagon observed in subjects undergoing prolonged exhaustive exercise.⁷¹ The exercise-induced increment in plasma glucagon appears dependent on the degree and duration of exercise, with some studies demonstrating no significant changes in plasma glucagon during exercise under normoglycemic conditions.⁷² Trained healthy male subjects exhibit increased hepatic glucose production following glucagon infusion, implying that exercise may be associated with the development of increased glucagon sensitivity.⁷³ Although prevention of exercise-induced rises in plasma glucagon by concomitant somatostatin infusion may result in mild hypoglycemia,⁷⁴ suppression of the rise in plasma glucagon by somatostatin infusion does not always prevent increased hepatic glucose production, likely because of the redundant compensatory mechanisms for maintaining normoglycemia.^74,75

The control of hepatic glucose production (HGP) is highly sensitive to the glucagon/insulin ratio, and secretion of these two hormones generally is regulated in a reciprocal manner. The effect of insulin to suppress HGP actually may be determined in part by the levels of circulating glucagon.⁷⁶ Following meal ingestion, nutrient absorption is associated with energy assimilation and suppression of glucagon secretion. Nevertheless, the integrated response of islet α cells is dependent in part on the nutrient composition of the meal and reflects positive and negative enteric-derived regulators of glucagon secretion. For example, ingestion of carbohydrate, especially glucose, is associated with a decline in plasma glucagon.⁶⁹ Nutrients also promote release of gut-derived peptides such as CCK and GIP, which stimulate glucagon release,^77,78 and GLP-1, which inhibits glucagon release.^79,80

Glucagon secretion is increased during times of stress, and “stress-induced hormones” such as cortisol, vasopressin, and β-endorphin increase glucagon secretion from the α cell. The classic stress hormones epinephrine and norepinephrine also stimulate glucagon secretion, and several mechanisms link the autonomic nervous system to increased secretion from the α cell. Epinephrine secreted from the adrenal medulla increases glucagon secretion in normoglycemic subjects. Furthermore, the pancreas receives innervation from both sympathetic and parasympathetic nerves, and stimulation of these autonomic inputs, all of which are activated by hypoglycemia, increases glucagon secretion.⁸¹ It is intriguing that mice with targeted inactivation of the pro-opiomelanocortin gene (and hence adrenocorticotropic hormone [ACTH]) exhibit a profound defect in the counterregulatory response to insulin-induced hypoglycemia, largely as the result of defective glucagon secretion. Although specific genetic defects in glucagon secretion have not been described, nondiabetic subjects with mutations in the MODY1/HNF-4α gene exhibit decreased arginine-stimulated glucagon secretion and reduced glucose suppression of plasma glucagon,⁸² raising the possibility that the HNF-4α transcription factor influences islet α cell function through a direct or indirect mechanism.

Hypoglycemia

Increasing evidence suggests that multiple complementary mechanisms activate α cell secretion during hypoglycemia. Analysis of the glycemic threshold for activation of counterregulatory mechanisms demonstrates that increased epinephrine and glucagon secretion constitute the initial hormonal responses to decreasing blood sugar. Furthermore, the glucose threshold for activation of counterregulatory responses is clearly higher than the threshold for triggering hypoglycemic symptoms in normal subjects.⁸³ Whether hypoglycemia itself directly stimulates glucagon secretion independent of autonomic input remains unclear. Nevertheless, hypoglycemia suppresses insulin (and β cell) secretion, which removes an important inhibitory influence on glucagon secretion. The finding that α cells express the GLUT1 transporter and glucokinase suggests that glucose transport does not appear to be a critical rate-limiting step for glucose metabolism in α cells and provides important insight into how α cells directly sense ambient changes in glucose concentration.^84,85 Hypoglycemia stimulates glutamate release from α cells, which, in turn, activates glucagon secretion in an autocrine manner.⁸⁶ Glutamate release from ventromedial hypothalamic neurons is also a key component of the counterregulatory response, as mice lacking the synaptic vesicular transporters (VGLUT2) exhibit fasting hypoglycemia and a defective glucagon response to hypoglycemia.⁸⁷ A role for gastrin in the stimulation of glucagon secretion has been proposed, and gastrin-/- mice exhibit defective glucagon secretion in response to insulin-induced hypoglycemia.⁸⁸

The role of circulating epinephrine in the autonomic response to hypoglycemia is well established. Epinephrine also directly stimulates glucagon secretion in normal human subjects, and stimulation of autonomic sympathetic nervous system innervation to the pancreas elicits an increase in α cell secretion.89,90 Furthermore, parasympathetic nerve stimulation or the neurotransmitter acetylcholine and the neuropeptide VIP all stimulate glucagon secretion. Studies using nerve transection or pharmacologic blockade have illustrated that these pathways exhibit some degree of functional redundancy, providing multiple backup mechanisms to ensure that the α cell responds appropriately to hypoglycemia. Nevertheless, the ganglionic blocker trimethaphan, which impairs autonomic transmission both in ganglia and in the adrenal, markedly attenuated the glucagon response to hypoglycemia in normal subjects,⁹¹ emphasizing the important link between autonomic activation and the glucagon counterregulatory response in vivo. In contrast however, pancreas transplantation restores the glucagon response to hypoglycemia, even in patients with severe autonomic neuropathy,⁹² illustrating the functional redundancy of mechanisms essential for the α cell response to hypoglycemia.

The identity of the specific glucose sensors that trigger the appropriate counterregulatory response to hypoglycemia remains under active investigation. Mice with genetic disruption of the Kir6.2 channel exhibit normal functional α cells, yet display a marked defect in the glucagon response to hypoglycemia.⁹³ GLUT2 expression in glial cells is essential for the glucagon response to hypoglycemia and the central nervous system, principally the ventromedial hypothalamus (VMH),⁸⁷ and the splanchnic region, specifically consisting of the portal system and liver, contains multiple glucose sensing systems. Selective perfusion of the portal venous system in rats suggests that the portal vein glucose concentration is a key determinant of the sympathoadrenal response to hypoglycemia; however, whether portal glucose sensors are directly or indirectly linked to control of islet glucagon secretion is not clear.⁹⁴ It is intriguing that injection of glucose into the portal vein activates glucose-sensitive neurons in the lateral hypothalamus and brain stem, suggesting the possibility of a portal-CNS glucoregulatory axis. The detection of a markedly defective counterregulatory response to hypoglycemia, including absent glucagon and catecholamine secretion, in a patient with hypothalamic sarcoidosis further emphasizes the importance of the hypothalamus in sensing glucose and triggering the release of counterregulatory hormones.⁹⁵ In contrast, analysis of patients after removal of craniopharyngiomas revealed selective impairment of counterregulatory sympathoadrenal activation but normal glucagon responses to hypoglycemia. Furthermore, human subjects with liver transplants and denervated livers exhibit increased levels of circulating glucagon and defective insulin suppression of glucagon, and human islet transplantation is associated with a defective glucagon response to hypoglycemia. These findings emphasize the importance of the autonomic nervous system and the CNS for the regulation of basal and hypoglycemia-stimulated glucagon secretion.

Glucagon Secretion and Diabetes

Glucagon levels are elevated in many patients with poorly controlled diabetes, and excess glucagon secretion in response to meal ingestion is an early event that is detectable within months of the onset of type 1 diabetes.⁹⁶ The correction of hyperglucagonemia, using somatostatin or insulin, reverses most metabolic derangements associated with insulin-deficient diabetes.⁹⁷ Hence there is considerable interest in development of glucagon antagonists for the treatment of diabetes. Moreover, genetic attenuation of glucagon receptor expression in diabetic rodents results in significant amelioration of experimental diabetes.^98,99 Despite the central importance of glucagon for control of glucose production, hyperglucagonemia alone, without insulin deficiency, does not significantly increase plasma glucose.¹⁰⁰ In the presence of adequate amounts of insulin, glucose production is suppressible despite glucagon excess, emphasizing the insulin:glucagon ratio, and not just the absolute level of glucagon, as a key determinant of glucose homeostasis.¹⁰¹ Nevertheless, hyperglucagonemia is associated with increased leucine oxidation and resting metabolic rate in subjects with type 1 diabetes, reemphasizing the importance of both insulin and glucagon in the catabolism associated with suboptimally treated diabetes.¹⁰²

The application of intensive insulin therapy to the management of patients with type 1 and type 2 diabetes is associated with an increased incidence of hypoglycemia and heightened awareness of the importance of counterregulatory mechanisms for maintaining normoglycemia. Several factors, including intensive insulin administration, hyperglycemia, and diminished autonomic stimulation of the α cell, inhibit glucagon release in insulin-treated patients with type 1 diabetes.⁸¹ Although the glucagon response to hypoglycemia is initially normal in patients with type 1 diabetes, this response frequently becomes impaired, increasing the susceptibility of patients to hypoglycemia.¹⁰³ Impaired counterregulation and hypoglycemia have been observed in some patients with a very short duration of diabetes, suggesting that frequent episodes of hypoglycemia represent an independent risk for development of an abnormal α cell response. The α cell dysfunction in type 1 diabetes is often selective, as the response to hypoglycemia may be absent, yet glucagon secretion may respond normally to arginine stimulation.¹⁰⁴ Further evidence emphasizing the importance of intra-islet hyperinsulinemia or other β cell–derived products in the suppression of glucagon secretion derives from observations that tolbutamide-infused subjects exhibit profound defects in the glucagon response to hypoglycemia.¹⁰⁵

It is important to note that restoration of normoglycemia for several months may decrease hypoglycemia unawareness in association with improvement in the glucagon response to hypoglycemia in some studies.¹⁰⁶ However, a dissociation between improvement in hypoglycemia unawareness and persistence of defective counterregulatory responses also has been observed.¹⁰⁷ Although not completely understood, antecedent hypoglycemia in type 1 diabetes can produce counterregulatory failure during subsequent episodes of prolonged moderate intensity exercise,¹⁰⁸ emphasizing the complex interrelationship between insulin-induced and exercise-associated counterregulation.

Multiple defects likely contribute to α cell dysfunction in patients with diabetes. Patients with some residual β cell function, as assessed by C-peptide stimulation, appear to be at decreased risk for hypoglycemia, in part because of preserved counterregulatory glucagon responses. Additional contributing factors may include subtle impairment of autonomic stimulation following repeated hypoglycemic episodes. Nevertheless, impaired epinephrine and glucagon responses to insulin-induced hypoglycemia have been observed after a single antecedent hypoglycemic episode in nondiabetic subjects,¹⁰⁹ although increased levels of cortisol alone do not acutely induce defects in hypoglycemic counterregulation. It is intriguing that oral administration of amino acids improves cognitive function and the glucagon response to hypoglycemia in both normal subjects and individuals with type 1 diabetes. Taken together, these observations emphasize the susceptibility of the normal α cell to episodes of hypoglycemia, ultimately leading to sustained defects in the glucagon response to hypoglycemia.

Glucagon Action

The importance of glucagon in the control of hepatic glucose metabolism provides a useful model for analysis of hormone action in metabolic pathways. Glucagon stimulates glucose production via activation of hepatic glycogenolysis and gluconeogenesis, and by inhibition of glycolysis. Following activation of the glucagon receptor, adenylyl cyclase activity is increased, leading to activation of protein kinase A and phosphorylase kinase, and increased rates of glycogenolysis via glycogen phosphorylase and inactivation of glycogen synthase.¹¹⁰ Glucagon action serves to modify the activity of enzymes important for glucose production via effects on specific kinases and phosphatases. Glucagon also modulates the expression of genes encoding enzymes of the glycolytic or gluconeogenic pathways¹¹¹ and regulates fatty acid metabolism via reduction of malonyl CoA and stimulation of fatty acid oxidation. The cAMP-dependent transcription factor is an important downstream mediator for glucagon action, acting in part through activation of the nuclear receptor coactivator PGC-1, and in part through suppression of peroxisome-proliferator–activated receptor-γ (PPAR-γ) agonist activity, which induces a metabolic program that results in the activation of hepatic gluconeogenesis.¹¹²

In adipocytes, glucagon increases cAMP and stimulates lipolysis, thereby providing free fatty acids as substrate for fat-burning tissues. Glucagon inhibits insulin-stimulated glucose transport in adipocytes through effects on insulin binding and via postreceptor mechanisms,¹¹³ and short-term glucagon administration reduces glucose uptake in hepatic and nonhepatic tissues.¹¹⁴ In the peripheral vascular system, glucagon functions as a vasodilator via effects on local vascular tone, and glucagon increases both cardiac output and heart rate, possibly via direct effects on the heart.¹¹⁵ Pharmacologic doses of glucagon increase renal blood flow, glomerular filtration rate, and urinary electrolyte excretion¹¹⁶; the kidney also exhibits significant gluconeogenic capacity and may account for up to 25% of systemic glucose production in humans.¹¹⁷ Although renal glucose output is markedly increased in subjects with diabetes, and hypoglycemia increases renal glucose output in association with increased release of counterregulatory hormones, the available evidence does not support an important role for glucagon in the control of renal glucose output.¹¹⁸

The actions of glucagon are transduced via activation of the glucagon receptor (Gcgr), a seven-transmembrane–spanning G protein–coupled receptor. The cloned receptor responds to glucagon with an increase in both intracellular cAMP and intracellular calcium.¹¹⁹ The human glucagon receptor gene has been localized to chromosome 17q25. Although activating mutations of the Gcgr have been generated by mutagenesis in vitro,¹²⁰ no constitutively active Gcgr mutations have been reported in human subjects. Although several population studies have described a Gly40Ser mutation linked to increased development of diabetes and/or hypertension, whether this mutation itself contributes to a diabetes predisposition or is associated with other genes that increase diabetes susceptibility in certain populations remains unclear.

The tissue distribution of Gcgr expression correlates well with studies localizing high-affinity glucagon-binding sites with Gcgr mRNA transcripts detected in liver, brain, adipocytes, heart, kidney, and islet β cells. Rat Gcgr mRNA transcripts also have been detected in spleen, thymus, adrenal gland, ovary, and testis—nonclassical target tissues where glucagon action remains poorly defined.¹²¹ Although glucagon action has been studied extensively in the liver and adipocytes, the precise biological importance of glucagon in the brain remains unclear. The brain stem is the principal site of central nervous system (CNS) proglucagon gene expression¹²²; however, PGDPs are transported from the brain stem along nerve fibers to multiple brain regions.¹²³ Consistent with these findings, glucagon binding sites are detected in multiple brain regions, and the glucagon receptor is expressed in cortex, cerebellum, hypothalamus, and brain stem; however, the specific biological action of glucagon in each of these CNS regions remains unclear. Although intracerebroventricular injection of glucagon in rodents causes hyperglycemia and increases sympathetic nervous system discharge, whether these findings are relevant to physiologic control of glucose homeostasis requires further analysis.

The β cell expresses receptors for glucagon, GIP, and GLP-1 that all are coupled to cAMP and stimulation of insulin secretion. The threshold for glucagon-stimulated cAMP accumulation in isolated β cells is ≈1 nM glucagon—higher than the concentrations required for cAMP stimulation by GLP-1 or GIP.¹²⁴ The physiologic importance of endogenous glucagon for β cell physiology in vivo remains unclear, given the direction of islet blood flow, the peripheral location of islet α cells in rodents, and the high concentrations of glucagon required to stimulate the β cell in vitro.

Insights into glucagon action derive from characterization of glucagon receptor knockout mice.17,125 Gcgr-/- mice exhibit marked elevations in levels of plasma glucagon, mild fasting hypoglycemia, increased pancreatic weight, α cell hyperplasia, and increased circulating levels of the PGDPs. The compensatory mechanisms sufficient for maintaining glucose production despite the complete absence of hepatic glucagon receptor signaling remain unknown. Similarly, the glucagon receptor–dependent signals regulating the number and secretory function of islet α cells have not yet been elucidated.

Both the liver and the kidney contribute to glucagon clearance from the circulation; however, these sites account for less than 50% of glucagon clearance, implicating additional tissues as sites for glucagon clearance or degradation.¹²⁶ Glucagon action is terminated via both extracellular and intracellular degradation pathways. Glucagon-degrading activity within hepatic endosomes has been attributed to cathepsins B and D in studies using cathepsin inhibitors,¹²⁷ and both glucagon and GLP-1 are substrates for the widely expressed membrane-bound neutral ectopeptidase (NEP) 24.11.¹²⁸ An endopeptidase activity has been described that cleaves 29 amino acid glucagon to “mini-glucagon,” which is also known as glucagon (19-29) in various tissues. The physiologic importance of mini-glucagon remains uncertain; whether glucagon (19-29) exerts physiologically relevant effects in the heart or β cell via a separate unique receptor remains unclear.

Pharmaceutical Use of Glucagon in Human Patients

Glucagon is employed as both a diagnostic and a therapeutic agent. Although historically useful for the diagnosis of pheochromocytoma, stimulation of catecholamine secretion in such patients may be dangerous, and hence the glucagon stimulation test is not widely used. Glucagon may also be employed as part of a diagnostic test in patients with hypoglycemia of unknown origin. Perhaps the most common therapeutic clinical application of glucagon is seen in the adjunctive management of severe hypoglycemia. Diabetic patients with hypoglycemia generally respond quickly with a rapid increase in blood glucose to intranasal, intramuscular, or subcutaneous glucagon.129,130 Glucagon is also used to inhibit gastrointestinal motility during radiologic investigations, and several studies have reported the efficacy of glucagon administration in small numbers of patients with bronchospasm, or symptomatic bradycardia.^131,132

Glucagon Excess and Deficiency

Glucagonomas presenting as solitary lesions or as part of a multiple endocrine neoplasia syndrome are most commonly detected in the pancreas and often are associated with significant elevations in the levels of circulating glucagon and PGDPs.¹³³ Rarely, extrapancreatic glucagon-producing tumors have been reported in sites including the kidney and ovary. Although many gut carcinoid tumors contain PGDP immunoreactivity, they are not generally associated with the development of a “glucagonoma syndrome.” Patients with glucagonoma generally present with a pathognomonic skin rash termed necrolytic migratory erythema, a detectable pancreatic mass, weight loss, glossitis, anemia, and some degree of glucose intolerance.¹³⁴ The clinical presentation can be variable, likely reflecting adaptation to the metabolic effects of tumor-secreted PGDPS and tumor-specific differences in posttranslational processing of proglucagon. Rarely, patients may present with glucagonomas and associated manifestations of intestinal hyperplasia, presumably due to release of GLP-2.¹³⁵ Treatment for benign glucagonomas usually involves surgical resection, whereas chemotherapy, attempted suppression of PGDP secretion and tumor growth with somatostatin analogues, or adjunctive radiotherapy may be indicated in patients with malignant disease. Experimental glucagonomas have been studied in rodents, and several intriguing phenotypes, including severe anorexia and reduction in islet size, remain poorly understood.^136,137 Several reports have described isolated cases of glucagon deficiency in infants with hypoglycemia; however, these cases are extremely rare. The molecular basis for the putative glucagon deficiency and whether congenital glucagon deficiency is compatible with survival remain unknown.

The Glucagon-like Peptides: GLP-1 and GLP-2

The proglucagon gene is expressed in the gastrointestinal tract in the stomach and both small and large intestine. The intestinal, brain, and pancreatic mammalian proglucagon mRNA transcripts are identical in structure; hence tissue-specific posttranslational processing underlies the liberation of the glucagon-like peptides in the brain and gastrointestinal tract. In contrast to the pancreas, much less is known about the control of intestinal GLP-1 and GLP-2 biosynthesis. The proglucagon gene islet transcription factors cdx-2/3, pax6 and members of the HNF3 (Foxa) family are also expressed in enteroendocrine cells and presumably regulate proglucagon gene transcription in both cell types. Neurogenin-3 has been identified as an essential upstream determinant of global enteroendocrine cell development. SEY^NEU mice with a dominant negative pax6 mutation exhibit a marked reduction in the levels of proglucagon mRNA transcripts in both small and large intestine. Hence, the pax6 gene is essential for both islet and enteroendocrine proglucagon gene transcription. Proglucagon biosynthesis in the gut is also regulated by nutrient intake,¹³⁸ with feeding and fiber-enriched diets increasing proglucagon gene expression in the proximal and distal intestine, respectively.

Glucagon-Like Peptide 1: Preclinical Studies

GLP-1, in both (7-37) and (7-36^amide) molecular forms, is liberated from proglucagon via posttranslational processing and is secreted from intestinal endocrine cells in a nutrient-dependent manner.¹³⁹ Although enteroendocrine L cells are distributed along the length of the entire gastrointestinal tract from the stomach to the rectum, the largest numbers of L cells are found in the terminal ileum and proximal colon. The rapid increase in plasma GLP-1 following food ingestion has fostered interest in the existence of a proximal-distal loop, whereby nutrients entering the duodenum and the proximal jejunum promote one or more endocrine and/or neural signals that activate GLP-1 secretion from the distal small bowel.¹⁴⁰ Studies in rats have identified GIP as one putative component of such a signaling system. The specific signaling mechanisms utilized by nutrients for stimulation of GLP-1 secretion in humans are not completely understood.

The regulation of GLP-1 bioactivity is dependent to a large extent on the rate of GLP-1 degradation and clearance in vivo. Both GLP-1 and GLP-2 contain an alanine residue at position 2, rendering these molecules substrates for enzymatic inactivation by dipeptidyl peptidase-4 (DPP-4). DPP-4 expressed locally in the intestine proximal to sites of GLP-1 synthesis cleaves circulating GLP-1 to yield GLP-1 (9-37/9-36^amide).¹⁴¹ Although GLP-1 (9-37/9-36^amide) displays weak binding affinity for the GLP-1 receptor and is theoretically a circulating antagonist of GLP-1 action, infusion of GLP-1 (9-36^amide) has modest insulin-independent effects on glucose tolerance and insulin secretion in human subjects.^142,143 Nevertheless, a substantial amount of total GLP-1 immunoreactivity circulates as GLP-1 (9-37/9-36^amide), and assays that do not distinguish between intact GLP-1 and cleaved GLP-1 (9-37/9-36^amide) will overestimate the circulating concentrations of bioactive GLP-1. More recent studies indicate that GLP-1 (9-36) amide may enhance cardiac function in preclinical studies^144,145 and may inhibit hepatic glucose production in obese human subjects.¹⁴⁶

Largely as a result of rapid DPP-4–mediated inactivation of GLP-1, the of circulating intact bioactive GLP-1 is very short, generally less than 1 minute. These findings have encouraged efforts at developing more potent long-acting GLP-1 analogues resistant to DPP-4–mediated inactivation. Alternatively, inhibitors of the DPP-4 enzyme have been identified that lower blood sugar in subjects with type 2 diabetes.147–149

GLP-1 exerts several complementary actions that control postprandial glycemic excursion, including stimulation of glucose-dependent insulin secretion and inhibition of glucagon secretion and gastric emptying (Fig. 9-2). GLP-1R agonists also regulate growth and development of the endocrine pancreas.¹⁵⁰ GLP-1 receptor activation promotes cell survival in β cells via increased levels of cAMP, leading to CREB activation, enhanced IRS-2 activity, and, ultimately, activation of Akt. These actions of GLP-1 are reflected by expansion of β cell mass and enhanced resistance to β cell injury demonstrated in multiple experimental models of diabetes in vivo. Moreover, Glp1r-/- mice exhibit modest defects in the formation of large islets and enhanced susceptibility to apoptotic injury.¹⁵¹ Of potential relevance to the therapeutic use of GLP-1 is the demonstration that primary human islet cultures exhibit improved glucose-dependent insulin secretion and enhanced survival following short-term exposure to GLP-1 in vitro.¹⁵² Similarly, use of the GLP-1R agonist exenatide is being explored in human subjects with type 1 diabetes following islet transplantation.¹⁵³

FIGURE 9-2 Biological activities of glucagon-like peptide (GLP)-1 and GLP-2. The various actions of GLP-1 (left panel) and GLP-2 (right panel) are shown.

Inhibition of gastric emptying accounts for a significant portion of the glucose-lowering actions of GLP-1, especially in patients with type 1 diabetes. Sensations of increased satiety and decreased appetite are common following GLP-1 infusion in human subjects,¹⁵⁴ and GLP-1 likely decreases appetite via central effects on the hypothalamic nuclei involved in feeding and via peripheral effects on gastric emptying. Although administration of GLP-1 and its analogues to rodents also results in decreased water intake, these actions do not appear to be clinically significant in human studies.

GLP-1 exerts its actions through activation of a G protein–coupled receptor that is structurally related to the glucagon/secretin receptor superfamily.¹⁵⁵ GLP-1 transduces its signal through both cAMP and calcium-dependent pathways. Although a potential candidate diabetes gene, the GLP-1 receptor (GLP-1R) localized to human 6p21 has not been associated with linkage to families with type 2 diabetes. Similarly, no GLP-1R mutations have been identified in individuals with diabetes, nor have activating mutations of GLP-1R been described in human subjects.

Evidence for the physiologic importance of GLP-1 in glucose control derives from studies using GLP-1 antagonists to inhibit GLP-1 action in vivo. Infusion of the GLP-1 antagonist exendin (9-39) into rats or humans increased blood glucose in association with decreased levels of glucose-stimulated insulin.156,157 Similarly, immunoneutralization of GLP-1 activity using GLP-1 antiserum increased both fasting- and meal-related glycemic excursions in baboons.¹⁵⁸ Furthermore, Glp1r-/- mice exhibit mild diabetes, with abnormalities in both fasting and postprandial glycemia and subnormal levels of glucose-stimulated insulin.¹⁵⁹

Therapeutic Use of GLP-1R Agonists for the Treatment of Patients with Type 2 Diabetes

Studies in human subjects have demonstrated the safety and efficacy of using GLP-1 to achieve glucose control in patients with type 2 diabetes. Native GLP-1 was administered via continuous subcutaneous infusion for 6 weeks to obese subjects with type 2 diabetes, resulting in highly significant improvement in glucose control, reduced glycosylated hemoglobin (HbA1c), improved insulin sensitivity, and a small but significant mean 1.9 kg weight loss.¹⁶⁰

Because the glucose-lowering effects of native GLP-1 administered by single subcutaneous administration are transient and are no longer evident 1 to 2 hours following peptide injection,¹⁶¹ therapeutic efforts were refocused on stable degradation-resistant GLP-1R agonists, including the lizard peptide exendin-4 (exenatide) and the human GLP-1 analogue liraglutide.¹⁶² As continuous enhancement of GLP-1 action for 24 hours a day appears superior for glucose control compared with shorter durations of peptide administration, current dogma supports the elevation of circulating levels of GLP-1 over a 24-hour period.

Exendin-4 (exenatide = synthetic exendin-4) is a 39 amino acid naturally occurring lizard GLP-1R agonist that exhibits 53% amino acid identity to mammalian GLP-1,163,164 yet binds to and activates the GLP-1 receptor. Furthermore, exendin-4 is highly resistant to the proteolytic activity of DPP-4 and exhibits a longer duration of action in vivo. Exenatide was evaluated in phase 3 trials, in combination with metformin, sulfonylurea agents, or both, for the treatment of patients with type 2 diabetes (T2DM), and its use led to significant lowering of HbA1c in a substantial percentage of treated subjects.^165–167 Subsequent studies confirmed the efficacy of adding exenatide to patients suboptimally controlled on a thiazolidinedione,¹⁶⁸ or as an alternative to insulin therapy.^169,170 Dose-limiting side effects of exenatide include nausea and vomiting. Most exenatide-treated subjects experience modest weight loss of 2 to 4 pounds over 6 months; however, about 20% of patients experience more substantial weight loss. Pancreatitis has been reported as a rare complication of therapy with GLP-1R agonists. A once-weekly form of exenatide is also in late-stage clinical development for the treatment of patients with type 2 diabetes.

Liraglutide is a fatty acid DPP-4–resistant human GLP-1 analogue administered as a once-daily therapeutic agent.¹⁷¹ Liraglutide reduces fasting and postprandial glycemia, inhibits gastric emptying, and reduces levels of circulating glucagon.¹⁷² Liraglutide has completed phase 3 clinical trials and is pending approval. Additional agents being evaluated in the clinic include a long-acting GLP-1 analogue suitable for once-weekly administration (taspoglutide) and a recombinant human albumin–GLP-1 protein (Albiglutide).

Oxyntomodulin is a 37 amino acid peptide that contains the sequence of glucagon together with an 8 amino acid carboxyterminal extension (see Fig. 9-1). Oxyntomodulin is capable of engaging both GLP-1 and glucagon receptors¹⁷³; however, a separate oxyntomodulin receptor has not yet been identified. Exogenous oxyntomodulin stimulates insulin secretion, lowers blood glucose, reduces food intake, and promotes weight loss in preclinical studies. Moreover, short-term oxyntomodulin administration is associated with enhanced satiety, weight loss, and increased energy expenditure in human subjects.^174,175 Hence there is active interest in evaluation of oxyntomodulin as a glucoregulatory and anorectic agent for the treatment of patients with T2DM and/or obesity.

Dipeptidyl Peptidase-4 Inhibitors

The challenges inherent in overcoming the rapid degradation of native GLP-1, together with the disadvantages and limitations of injectable peptides, have fostered complementary efforts at enhancing endogenous levels of native GLP-1 through the use of dipeptidyl peptidase-4 inhibitors.147,149 DPP-4 inhibitors preserve circulating levels of bioactive GLP-1 and GIP and exert significant glucose-lowering actions in rodents and human subjects. The DPP-4 gene is essential for incretin degradation and glucose homeostasis in mice, as genetic elimination of DPP-4 results in increased levels of intact plasma incretins, enhanced insulin secretion, and reduced glycemic excursion.¹⁷⁶ The observation that DPP-4 inhibitors produced significant glucose-lowering effects in four 12-week human studies¹⁷⁷ has fostered development of several highly selective orally available DPP-4 inhibitors suitable for once-daily administration. Sitagliptin, generally administered as a 100 mg once-daily tablet, is the first widely approved DPP-4 inhibitor¹⁴⁸; it lowers blood glucose when used as monotherapy or as add-on therapy to other oral antidiabetic agents in a broad spectrum of patients with T2DM.^178–180 The most compelling use of DPP-4 inhibitors in subjects with T2DM is seen in early initial combination therapy with metformin,¹⁸¹ as metformin appears to independently increase the circulating levels of GLP-1. DPP-4 inhibitors are well tolerated, are weight neutral, and are associated with a low risk for hypoglycemia. A second highly selective DPP-4 inhibitor, alogliptin, has also completed phase 3 clinical trials in subjects with T2DM. As DPP-4 regulates immune function and controls the degradation of a broad spectrum of endocrine and regulatory peptides, the long-term safety of DPP-4 inhibitors, which specifically target the kinase activity of the molecule, remains under scrutiny. Studies of the mechanism of action of DPP-4 inhibitors in vivo reveal that the glucoregulatory actions of DPP-4 inhibitors are abolished in mice with combined genetic disruption of both GIP and GLP-1 receptors.¹⁸² The overlapping and contrasting actions of GLP-1R agonists versus DPP-4 inhibitors are illustrated in Table 9-1.

Table 9-1

GLP-1 Receptor Agonists Versus DPP-4 Inhibitors

	GLP-1R Agonists	DPP-4 Inhibitors
Administration	Injection	Orally available
GLP-1 concentrations	Pharmacological	Physiological
Mechanisms of action	GLP-1	GLP-1 + GIP
Activation of portal glucose
Sensor	No	Yes
↑ Insulin secretion	+++	+
↓ Glucagon secretion	++	++
Gastric emptying	Inhibited	+/−
Weight loss	Yes	No
Expansion of β-cell mass
In preclinical studies	Yes	Yes
Nausea and vomiting	Yes	No
Potential immunogenicity	Yes	No

Adapted from Cell Metab 3(3):153-165, 2006.

Glucagon-Like Peptide-2

GLP-2 is co-secreted with GLP-1 from intestinal endocrine cells and is trophic to the mucosal epithelium in both the small and large intestine (see Fig. 9-2).¹⁸³ The presence of an alanine residue at position 2 (see Fig. 9-1) predicted that GLP-2, like GLP-1, would be a substrate for inactivation by DPP-4, and significant amounts of biologically inactive GLP-2 (3-33) have been demonstrated in both rodent and human plasma.¹⁸⁴ Similarly, analogues of GLP-2 designed to resist DPP-4–mediated cleavage are biologically more potent in vivo. Intravenous administration of GLP-2 stimulates intestinal glucose transport within 30 minutes, and a single injection of GLP-2 rapidly enhances barrier function in the mouse gut, demonstrating that GLP-2 exerts acute actions independent of bowel growth in the gastrointestinal tract.

The importance of GLP-2 for maintenance of the intestinal villous epithelium is illustrated by studies of rats receiving total parenteral nutrition (TPN), as co-infusion of GLP-2 and TPN prevented mucosal villous hypoplasia in the small intestine.¹⁸⁵ The physiologic importance of GLP-2 for the physiology or growth of the normal mucosal epithelium has been examined in mice administered GLP-2 (3-33), a partial GLP-2 antagonist. Refeeding of fasted mice results in enhanced mucosal growth; however, coadministration of GLP-2 (3-33) significantly blunted the adaptive response to enteral nutrients.¹⁸⁶ Similarly, treatment of diabetic rats with immunoneutralizing GLP-2 antisera significantly reduces adaptive mucosal growth in the small bowel.¹⁸⁷

A G protein–linked GLP-2 receptor, related in sequence to the glucagon and GLP-1 receptors, has been isolated ¹⁸⁸ and has been localized to human enteroendocrine cells, myofibroblasts, and enteric neurons. GLP-2 promotes regeneration and prevents apoptosis in experimental models of intestinal resection and inflammation, including small bowel enteritis, chemotherapy-induced mucosal injury, and ischemic damage.^189–191 Furthermore, the antiapoptotic actions of GLP-2 have been demonstrated in primary rat and mouse hippocampal cell cultures and in heterologous cells expressing a transfected GLP-2 receptor. Administration of native GLP-2 twice daily to human subjects with short bowel syndrome resulted in improved energy absorption, decreased fluid loss, and significant weight gain, in association with increased crypt plus villus height in mucosal biopsy specimens.^192,193 A degradation-resistant stable human GLP-2 analogue, teduglutide, is currently being evaluated in clinical trials of patients with short bowel syndrome and inflammatory bowel disease.

In summary, the importance of pancreatic derived glucagon in control of glucose homeostasis encompasses the critical role of glucagon as a primary defense against hypoglycemia. Recently, interest has been renewed in strategies designed to block inappropriate glucagon action for the treatment of patients with type 2 diabetes. Furthermore, the delineation of multiple biological actions of intestinally derived GLP-1, GLP-2, and oxyntomodulin and the isolation of distinct receptors for these peptides have greatly expanded our understanding of the biological relevance and actions of the proglucagon-derived peptides. It seems reasonable to postulate that additional, as yet undescribed, actions of these PGDPs remain to be elucidated. Finally, the therapeutic utility of GLP-1R agonists, DPP-4 inhibitors, and GLP-2R agonists has rekindled interest in enhanced understanding of the biology of these peptides, with greater emphasis on understanding the long-term safety of chronic PGDP administration in human subjects.