Contents
Contributors
Search

Glucagon is a 29 amino acid polypeptide (molecular weight of 3485 daltons) (Figure 1) which was discovered as a "contaminant" hyperglycemic factor in pancreatic extracts by Kimball and Murlin (1) in 1923 and finally sequenced by Bromer and Behrens (2) in the late 1950s. Studies of its mechanism in the 1960s by Sutherland et al (3) led to the discovery of the second messenger cyclic adenosine monophosphate (cAMP) for which the Nobel Prize was awarded. Full appreciation of importance of glucagon for normal fuel homeostasis in humans and abnormalities in patients with diabetes mellitus did not come until the 1970s when a specific radioimmunoassay was developed and when the availability of somatostatin, an inhibitor of glucagon secretion, permitted investigation of its lack under various experimental conditions (4).

Figure 1. Amino acid sequence of glucagon

BIOSYNTHESIS

Glucagon is synthesized in and secreted from A cells of pancreatic islets. Normally, these cells constitute approximately 15% to 20% of the total islet cell mass. In most species, A cells are located at the periphery of islets juxtaposed to both B cells, which secrete insulin, and D cells, which secrete somatostatin which can inhibit both insulin and glucagon secretion (5). In humans, however, A cells are scattered throughout the islet. Granules in A cells containing glucagon differ in ultrastructure from those in B and D cells containing respectively insulin and somatostatin in having an electron-dense core and no halo (Figure 2).

Glucagon is synthesized initially as a 160 amino acid prohormone (proglucagon) of approximately 12,000 d whose gene is encoded on chromosome 2. Proglucagon ultimately undergoes cleavage into four peptides (Figure 3) (6,7). The whole process takes about 90 minutes. All of these peptides are immunoreactive, but only the 3485-d molecule is biologically active. L-cells of the small intestine synthesize an identical proglucagon molecule but different processing results in the formation of different polypeptides, of which glucagon-like peptide 1 and 2 are probably of most physiologic importance.

Figure 3. Human proglucagon with differential processing in pancreatic alpha cells and intestinal L cells. In the alpha cells, the major products released into plasma are glicentin related polypeptide, glucagon, a minor and a major proglucagon fragment whereas in the intestinal L cells, the major products released are glicentin, GLI-1 (glucagon-like peptide-1) and GLP-2 (glucagon-like peptide-2). (Copyright 2001 by McGraw-Hill Companies. Reproduced with permission of McGraw-Hill Companies from Masharani U and Karam JH: Pancreatic hormones and diabetes mellitus. In: Basic & Clinical Endocrinology. Greenspan FS and Gardner DG eds. New York: McGraw-Hill, 2001, pp. 623-698.)

PLASMA GLUCAGON

Plasma immunoreactive glucagon concentration varies considerably from individual to individual. The main factors responsible for this variation are the specificity of the antiserum used in the immunoassay and the relative proportion of the total immunoreactivity accounted for by the 3485-d molecule (4,8,9).

Normally, in humans and most other mammalian species, arterial and peripheral venous plasma immunoreactive glucagon concentrations range between 25 and 150 pg·ml-1 (1.0-5.0 x 10-8M) after a 12- to 16-hour fast. Portal venous levels can average 1.5 to 3.0 times those present in arterial blood because of extraction of glucagon by the liver (10-17). As with other peptide hormones, circulating glucagon immunoreactivity is heterogeneous (9). By using chromatography, four immunoreactive species with apparent molecular weights of >40,000, 9000, 3500, and 2000 have been found (Figure 4) (Table 1). There is considerable individual and species variation in the proportions of each component found in plasma (12,13). In early studies, the 3500-d species usually constituted only about 25% of total plasma glucagon immunoreactivity due to nonspecificity of assay. With subsequent improvements and extraction procedures now available, it probably accounts for 90-95%. The 9000-d molecule, which in some assays has similar immunoreactivity but substantially less bioactivity than the 3500-d molecule, can be converted by trypsin to a smaller immunoreactive peptide of approximately 3500 d (18); it is thought to represent the biosynthetic precursor of glucagon found in the pancreas, which is also convertible to glucagon by trypsin (19). Increased amounts of the 9000-d molecule are found in the plasma of patients with the glucagonoma syndrome (20), with renal failure (14), and hepatocellular damage or carcinoma of the pancreas (8). The 2000-d molecule probably represents an inactive degradation product of glucagon.

Figure 4. Immunoheterogeneity of human plasma glucagon. (Copyright 1998 by Springer-Verlag. Reproduced with permission of Springer-Verlag from Gerich J and Cryer P. Counterregulatory hormones: molecular, biochemical, and physiologic aspects. In: Principles of Perinatal-Neonatal Metabolism. 2nd edition. Cowett R, ed. New York: Springer-Verlag, 1998, pp. 155-180.)

The heterogeneity of plasma glucagon has complicated the interpretation of in vivo studies of glucagon secretion and metabolism. Changes in plasma glucagon immunoreactivity during stimulation or suppression of A cell secretion are due almost exclusively to changes in the 3500-d fraction (18,21,22). Although the overall distribution of plasma glucagon is not altered in diabetes and most other pathologic conditions in which the study of A cell function might be of interest (9), the relative contribution of the fractions can vary considerably among individuals. Thus comparisons, based on absolute levels of total plasma glucagon immunoreactivity using early assays, may have been misleading.

The pancreatic content of glucagon varies considerably among species; the human pancreas contains approximately 700 to 1000 µg of glucagon, roughly 1-2 x 1017 molecules. For reference, the normal human pancreas contains about ~150 U of insulin (23,24) or ~5000 µg; this translates to ~6-8 x 1017 molecules. Thus the normal human pancreas has about 1/3 to 1/10 the number of glucagon molecules as insulin molecules.

Glucagon, stored within A cells in distinctive granules, is secreted by a process called emiocytosis (25), which involves migration of secretory granules to the periphery of cells, fusion of granules with the plasma membranae, and extrusion of granule contents into the extracellular space.

Like insulin, secretion of glucagon involves A cell substrate metabolism and consequent signals which affect cellular potassium and calcium channels and cAMP levels as well as protein kinase A and C (Figures 5 and 6) (26,27).

Figure 5. Schematic for stimulators of glucagon release

Figure 6. Schematic for inhibition of glucagon secretion

Most substrates (glucose, free fatty acids and ketone bodies) except certain amino acids suppress glucagon secretion (26). Inhibition of the metabolism of these substrates prevents the inhibition (26) suggesting in contrast to insulin secretion by beta cells that generation of ATP inhibits secretion. While this is consistent with the reciprocal roles of insulin and glucagon in glucose homeostasis, a definitive explanation for this difference remains to be elucidated.

Alpha cells contain ATP-sensitive potassium channels as well as sulfonylurea, adrenergic, insulin and somatostatin receptors (28). Sulfonylurea receptors and ATP-sensitive potassium channels are associated with both plasma membranes and secretory granule membranes (29). Sulfonylureas stimulate glucagon release under appropriate conditions (30). This effect is dependent on protein kinase C, mimicked by inhibitors of mitochondrial ATP-sensitive potassium channels and inhibited by K+ channel openers (diazoxide) (31).

It has been proposed (31) that the effect of sulfonylureas on alpha cell granules involves alterations in granule pH which renders them more competent for emiocytosis. Since metabolism of substrates would be expected to generate ATP and thus inhibit ATP-sensitive potassium channels like sulfonylureas, it is difficult to reconcile these observations into a consistent molecular mechanism for acute regulation of glucagon secretion. However, it has been proposed that a decrease in the intra-alpha cell ATP/ADP ratio activates adenylate cyclase and the resultant increase in cyclic AMP stimulates protein kinase A which causes opening of calcium channels and an increase in intra-alpha cell calcium which triggers glucagon release (Figure 6).

In vivo secretion of glucagon is the net result of the influence of substrate, neural, ionic, hormonal, and local factors on islet A cell function. The plasma concentration of glucagon depends on the balance between rates of secretion and degradation and also on the sampling site (e.g. peripheral venous versus portal venous). Basal (nonstimulated) secretion rates of glucagon can be estimated from data on portal venous-arterial differences and portal venous plasma flow rates. Secretion rates of glucagon may also be estimated on the basis of the clearance of glucagon under steady-state conditions; such estimation yields a value of approximately 1400 pg·kg^-1·min^-1 in humans (32). It should be pointed out, however, that these values underestimate secretion of glucagon and merely represent posthepatic delivery of glucagon. From what is known of the pancreatic content of glucagon and secretory rates of glucagon, it can be estimated that at least 25%, and probably more, of the pancreatic content of glucagon is secreted each day.

GLUCAGON CATABOLISM

In normal humans, the metabolic clearance rate of glucagon is independent of the prevailing plasma glucagon level. Estimates range between 7 and 14 ml·kg^-1min^-1 (32,33). Normal rates occur in patients with diabetes (33) or liver disease (34), whereas decreases have been found in renal failure (35) and starvation (32). Thus the liver and kidney seem to be the major sites of glucagon catabolism, but the relative contribution of each remains unclear (34,35).

Early reports suggested that the liver was not a major site of glucagon degradation (10,14,16). These observations may, however, be explained if the heterogeneity of circulating glucagon immunoreactivity is taken into account. When portal venous and peripheral venous plasma is subjected to gel filtration, it seems that the liver does not appreciably extract the biologically inactive 9000- and >40,000-d plasma glucagon immunoreactivity (14). Thus, the portal-peripheral gradient of glucagon immunoreactivity is almost totally accounted for by extraction of the biologically active 3500-d molecule; this averages approximately 60% and results in a portal-peripheral gradient of 2.5 to 3 for the biologically active molecule.

It has long been known that the kidney is capable of degrading glucagon. Arteriovenous gradients across the kidney in normal animals infused with glucagon indicate extraction of 23% to 39% of the presented glucagon (12,36,37). Because less than 2% of the extracted hormone appears in urine and because nonfiltering kidneys continue to extract appreciable amounts of glucagon (37), it seems that both tubular reabsorption and postglomerular capillary tubular uptake precede renal parenchymal degradation of glucagon. The hyperglucagonemia found in patients with chronic renal failure is due primarily to decreased clearance of the 9000-d molecule and cannot be accounted for by increased secretion of glucagon (3500-d molecule) or its decreased catabolism (38). Bilateral nephrectomy decreases the glucagon metabolic clearance rate of 3500-d glucagon approximately 30% (12). Consequently, liver and kidney can account for 80% to 90% of the metabolic clearance of the biologically active glucagon fraction of plasma glucagon immunoreactivity.

REGULATION OF GLUCAGON SECRETION

Glucose is the most important physiologic regular of glucagon secretion. Hyperglycemia decreases and hypoglycemia increases glucagon secretion (26). In vitro studies, such as those using the isolated perfused pancreas in which most variables operative in vivo can be controlled, indicate that the A cell is as exquisitely sensitive to changes in the ambient extracellular glucose concentration as is the B cell (39) (Figure 7); thus, glucose suppresses basal and stimulated glucagon release at concentrations as low as 5 mM glucose (90 mg·dl^-1) (Figure 8). To some extent the inhibition of glucagon secretion is dependent on concomitant stimulation of insulin release. In vivo, a decrease in plasma glucose of 1 to 2 mM increases plasma glucagon (40).

Figure 7. Effect of glucose on glucagon and insulin release from the in vitro perfused rat pancreas. (Copyright 1974 by J Clin Invest. Republished with permission of J Clin Invest from Gerich et al: Characterization of the effects of arginine and glucose on glucagon and insulin release from the perfused rat pancreas. J Clin Invest 54:833-841, 1974.)

Figure 8. Dose-response relationships for effect of glucose to suppress glucagon and stimulate insulin release from the perfused rat pancreas. (Copyright 1974 by J Clin Invest. Republished with permission of J Clin Invest from Gerich et al: Characterization of the effects of arginine and glucose on glucagon and insulin release from the perfused rat pancreas. J Clin Invest 54:833-841, 1974.)

Other substrates also influence glucagon secretion. Various amino acids stimulate A cell release of glucagon (41), while free fatty acids (42) and ketone bodies (43) suppress glucagon secretion (Figure 9). Amino acid stimulation of glucagon release may be important in preventing hypoglycemia, which might otherwise occur because of insulin release accompanying ingestion of a noncarbohydrate meal. Suppression of glucagon secretion by free fatty acids and ketone bodies may be part of a negative feedback system regulating ketogenesis.

Figure 9. Effect of elevation of plasma free fatty acids on plasma glucagon levels in normal human volunteers. (Copyright 1974 by J Clin Invest. Republished with permission of J Clin Invest from Gerich et al: Effects of alterations of plasma free fatty acid levels on pancreatic glucagon secretion in man. J Clin Invest 53:1284-1289, 1974.)

The islets of Langerhans are richly innervated. Like insulin release, glucagon secretion is influenced by both sympathetic and parasympathetic nervous sytems; epinephrine, norepinephrine, and acetylcholine (44,45), and electrical stimulation of mixed pancreatic, splanchnic, and vagus nerves augment glucagon release (46,47). Both A and B cell secretion are influenced in the same direction by parasympathetic (i.e. increase), ß-adrenergic (i.e increase), and a-adrenergic (i.e. decrease) mechanisms. The observation (Figure 10) that glucagon secretion is increased by epinephrine while insulin release is simultaneously decreased can best be explained by postulating that the A cell contains a preponderance of ß-adrenergic receptors, while the B cell contains a preponderance of a-adrenergic receptors. Neural input to the A cell is probably important in modulating the increases in plasma glucagon observed during stress and perhaps also after mixed meals.

Figure 10. Effect of infusion of epinephrine on plasma glucagon and insulin levels in normal volunteers. (Copyright 1973 by The Endocrine Society. Reproduced with permission of The Endocrine Society from Gerich JE, Karam JH, Forsham PH: Stimulation of glucagon secretion by epinephrine in man. J Clin Endocrinol Metab 37:479-481, 1973.)

A variety of hormones have been reported to alter A cell function; epinephrine, gastrin, pancreozymin, vasoactive interspinal peptide, and gastric inhibitory polypeptide increase glucagon release (48,49), while secretin apparently suppresses glucagon secretion (50). Whether these represent true physiologic interactions or merely pharmacologic effects is unclear. Hyperglucagonemia, relative or absolute, has been found in states of growth hormone (51), cortisol (52), and thyroid hormone excess (53). Conceivably, this might play a role in the associated abnormalities of carbohydrate and lipid metabolism.

Alterations in nutrition also influence A cell function. Acute ingestion of pure or high carbohydrate meals suppresses glucagon release, whereas pure or high protein-containing meals stimulate glucagon release. Concomitant changes in plasma glucose and amino acid levels are probably responsible for these changes. Prolonged (i.e. weeks or days) alterations in diet also alter A cell function. During total starvation, there is an acute increase in plasma glucagon lasting 1 to 2 days, probably as a result of increased secretion (54). Prolonged ingestion of high-carbohydrate or isocaloric high-fat diet decreases basal and meal-stimulated plasma glucagon levels (55). Conversely, low-carbohydrate diets or high-protein diets increase basal and stimulated glucagon secretion (55). In obesity, increased plasma glucagon responses have been reported (56).

Hypoglycemia stimulates glucagon secretion through both intraislet and central nervous system mediated autonomic signals (57). Within the islets low glucose concentrations increase A-cell glucagon secretion directly and, by reducing B-cell secretion, decrease tonic A-cell inhibition by insulin. Autonomic adrenergic (i.e. norepinephrine), cholinergic, and peptidergic neural and adrenomedullary hormonal (epinephrine) signals, triggered by hypoglycemia, may also contribute.