Diabetes and the Alpha Cell

In the post-absorptive period (overnight fast), a quasisteady state exists in which the rate of glucose release into plasma approximates its removal. Under these conditions, most of glucose removal from plasma is independent of insulin (i.e. it is largely determined by obligatory tissue demands). For example, glucose uptake by the brain, the formed elements of the blood, the renal medulla, and the intestine, all of which do not require insulin, account for 70–80% of glucose disposal.¹


In the post-absorptive period (overnight fast), a quasisteady state exists in which the rate of glucose release into plasma approximates its removal. Under these conditions, most of glucose removal from plasma is independent of insulin (i.e. it is largely determined by obligatory tissue demands). For example, glucose uptake by the brain, the formed elements of the blood, the renal medulla, and the intestine, all of which do not require insulin, account for 70–80% of glucose disposal.¹

In contrast, the release of glucose into the circulation is mainly regulated by insulin, glucagon, and catecholamines. The liver and kidneys account for ~80% and ~20% of glucose release, respectively.³ Insulin restrains hepatic and renal glucose release, while glucagon and catecholamines promote glucose release by the liver and kidneys, respectively. It is thus clear that in the post-absorptive period, the release of glucose into the circulation is largely the result of the actions of insulin and glucagon and is the primary means by which plasma glucose concentrations are regulated.

In the postprandial period (e.g. the four- to six-hour period following meal ingestion), insulin-mediated tissue uptake of glucose assumes a more important role than in the post-absorptive period; however, regulation of glucose release into the circulation is still the most important factor.⁴ Normally after meal ingestion, glucose release by the liver from endogenous sources (glycogenolysis and gluconeogenesis) is suppressed, while glucose release into the circulation from carbohydrates in the meal is largely dependent on first-pass hepatic extraction. In addition, the liver takes up glucose from the systemic circulation, such that after meal ingestion, total glucose uptake by the liver is approximately the same as muscle glucose uptake.From the foregoing, it is clear that, as in the postabsorptive period, the liver plays an important role in the postprandial period. Although glucose transport into hepatocytes does not require insulin, the net uptake of glucose by the liver is dependent on insulin and glucagon since insulin suppresses glycogen breakdown while glucagon promotes glycogen breakdown.¹ After meal ingestion, insulin release increases and glucagon release decreases, resulting in a marked increase in the plasma insulin:glucagon ratio. This ratio is inversely related to the rate of glucose release into the circulation after meal ingestion.⁵

The major determinants of postprandial secretion of insulin and glucagon are the plasma glucose level, an individual’s insulin sensitivity,⁶ and incretin hormones (e.g. glucagon-like peptide-1 (GLP-1) and gastric inhibitory polypeptide (GIP), predominantly the former).7 These factors reciprocally affect insulin and glucagon secretion: namely, increases in plasma glucose and GLP-1 stimulate insulin and suppress glucagon secretion.

Given the importance of insulin and glucagon in both fasting and postprandial glucose homeostasis, abnormalities in the release of and tissue sensitivity to these hormones would be expected to adversely affect glucose tolerance. Studies performed in the second half of the 20th century established that, indeed, this was the case. However, during the last two decades, considerable emphasis has been placed on impaired β- cell function and insulin resistance as the key factors in the pathogenesis of type 2 diabetes, while the important role of α-cell dysfunction seems to have been forgotten.

Recently, however, the development of therapeutic agents that mimic the action of GLP-1 (e.g. GLP-1 analogs) or increase endogenous GLP-1 levels (e.g. dipeptidyl peptidase (DPP)-4 inhibitors) has led to renewed interest in α-cell dysfunction because these agents act in part by suppressing glucagon secretion.^7,8 It is the purpose of this review to summarize the evidence for α-cell dysfunction in type 2 diabetes and the potential beneficial effects of the suppression of glucagon secretion on glycemic control. Pancreatic Islet Cell Dysfunction in Type 2 Diabetes
It is now well established that in type 2 diabetes, insulin secretion is decreased. In the United Kingdom Prospective Diabetes Study (UKPDS), homeostatic model assessment (HOMA) indicated that β-cell function was reduced by ~50% and subsequently progressively deteriorated.⁹ Decreased insulin secretion has also been demonstrated in people with impaired glucose tolerance (IGT), as well as in first-degree relatives of people with type 2 diabetes who still have normal glucose tolerance.^5,10 Although decreases in β-cell mass have been found in people with IGT and type 2 diabetes,¹¹ functional defects also exist, as demonstrated by in vitro studies of islets harvested from the pancreata of people with type 2 diabetes.¹²

These changes in β-cell function and mass are accompanied by changes in α-cell function and mass.^13,14 In the post-absorptive period, both plasma insulin and glucagon levels are inappropriate for the prevailing plasma glucose levels.Thus, although plasma insulin levels may be increased in patients with type 2 diabetes compared with levels in normoglycemic nondiabetic subjects, the plasma insulin levels of the latter become two- to three-fold greater than those of patients with diabetes when plasma glucose levels are increased to match those of patients with diabetes).¹⁵ Similarly, plasma glucagon levels in patients with type 2 diabetes have repeatedly been demonstrated to be either comparable with those of normoglycemic non-diabetic subjects despite hyperglycemia, or to be frankly increased.^13,16

Other functional abnormalities of the α-cell include:¹⁶

excessive increases in glucagon secretion during amino acid infusions;¹⁷
1. reduced plasma glucagon stimulation during insulininduced hypoglycemia;¹⁸ and
2. lack of appropriate suppression of glucagon secretion after meal ingestion.⁴

This latter abnormality has been demonstrated to occur in people with IGT who also have decreased insulin responses.⁵ The resultant marked reduction in plasma insulin:glucagon molar ratios are inversely correlated with postprandial glucose release into the circulation and postprandial hyperglycemia.⁵Causes of α- Cell Dysfunction
The etiology of α-cell dysfunction in type 2 diabetes is complex and still poorly understood. In type 1 diabetes, hyperglucagonemia may merely be a consequence of insulin deficiency; however, this does not appear to be the case in type 2 diabetes. First, absolute increases in pancreatic α-cell mass have been reported.^12-14 Second, fasting hyperglucagonemia often coexists with fasting hyperinsulinemia.¹⁶ Third, infusion of insulin to produce supraphysiologic plasma insulin levels does not correct α-cell dysfunction.¹⁷ Consequently, it has been proposed that a similar fundamental ‘blindness’ to glucose by the islet β-cell may be operative in α-cells of patients with type 2 diabetes.^13,18

Altered islet architecture, disrupting neural regulation of insulin secretion, could also be involved. Another factor that is probably involved is a decreased incretin effect. Patients with IGT and those with type 2 diabetes have reduced increases in plasma GLP-1 concentrations after meal ingestion.¹⁹ Since GLP-1 suppresses glucagon secretion, this may contribute to the lack of appropriate postprandial suppression of glucagon release. Finally, α-cells have insulin receptors, and insulin suppresses glucagon secretion. Therefore, α-cells may be resistant to insulin, as are other tissues in individuals with type 2 diabetes.

Contribution of α- Cell Dysfunction to Fasting and Postprandial Hyperglycemia

Fasting Hyperglycemia
That a mere deficiency of insulin could not explain the complete development and maintenance of fasting hyperglycemia was first demonstrated more than 30 years ago20 (see Figure 1). After an intravenous infusion of insulin that had been maintaining type 1 diabetic subjects in a euglycemic state was stopped, plasma glucose levels increased only to ~150mg/dL when somatostatin was infused to suppress glucagon secretion. In contrast, when saline was infused instead of somatostatin, plasma glucagon increased and plasma glucose concentrations rose to nearly 300mg/dL after stopping the insulin infusion. When exogenous glucagon was infused along with somatostatin to reproduce the hyperglucagonemia observed in the saline experiments, hyperglycemia and hyperketonemia occurred at levels similar to those observed in the saline experiments. Furthermore, when somatostatin was infused six hours after stopping the insulin infusion, plasma glucose levels decreased from ~300mg/dL to ~150mg/dL over six hours despite continued insulin deprivation. Recently, similar attenuation of fasting hyperglycemia has been demonstrated in patients with type 1 during infusion of GLP-1, suppressing glucagon secretion.²¹ In these experiments, plasma glucose concentrations decreased from ~250mg/dL to 180mg/dL during a four-hour infusion of GLP-1.

Subsequent studies using somatostatin²² and GLP-123 to reduce hyperglucagonemia in patients with type 2 diabetes have confirmed the important contribution of glucagon to the maintenance of fasting hyperglycemia. After a two-hour infusion of somatostatin,²² plasma glucose levels decreased from ~250mg/dL to ~190mg/dL, and hepatic glucose output decreased to normal values. After a four-hour infusion of GLP-1,23 plasma glucose levels decreased from ~230mg/dL to <100mg/dL.Postprandial Hyperglycemia
In the 1970s, studies in patients with type 1 diabetes using somatostatin to suppress postprandial hyperglucagonemia demonstrated an improvement in postprandial hyperglycemia with and without insulin administration.24 Subsequent experiments in patients with type 2 diabetes have provided strong evidence for a role of α-cell dysfunction in determining postprandial hyperglycemia.^25-28

Mitrakou et al.5 demonstrated that the magnitude of postprandial hyperglycemia in people with IGT was primarily related to increased rates of glucose delivery into the circulation and that this was inversely related to the postprandial plasma insulin:glucagon molar ratio.

Recently, Henkel et al.25 confirmed the findings of Mitrakou et al. using multivariate analysis of data in patients with IGT and patients with type 2 diabetes, indicating that plasma insulin and glucagon responses were significant independent determinants of postprandial hyperglycemia.

In an elegant series of acute interventional experiments, Shah et al.26-27 and Dineen et al.28 demonstrated that the characteristic lack of suppression of glucagon secretion after meal ingestion contributed significantly to postprandial hyperglycemia. Specifically, these investigators showed that lack of postprandial suppression of glucagon secretion adversely affected glucose tolerance even in the face of ‘normal’ plasma insulin profiles and much more so when accompanied by ‘diabetic’ plasma insulin profiles.²⁷

Suppression of Hyperglucagonemia as a Therapeutic Target
From the foregoing, it would appear that reducing hyperglucagonemia might have beneficial effects on glycemic control in patients with diabetes. Nearly 30 years ago, a proof-of-concept experiment showed that a three-day infusion of somatostatin substantially reduced hyperglycemia throughout each day in both type 1 and type 2 diabetes²⁹ (see Figure 2) while reducing insulin requirements by about 50%. However, subsequent attempts to develop clinically useful selective glucagon-suppressing somatostatin analogs or glucagon receptor antagonists have not been successful.

Recently, however, the development of GLP-1 analogs (e.g. exenatide, liraglutide) and DPP-4 inhibitors (e.g. vildagliptin, sitagliptin, saxagliptin), which increase endogenous active GLP-1 and GIP, have provided a means to exploit the advantages of suppressing glucagon secretion.⁷ It has been estimated, for example, that ~50% of the reduction in postprandial glucose production by exenatide is due to its suppression of glucagon secretion.8 In patients with type 2 diabetes, these agents appear to be as efficacious as and complementary to sulfonylureas, metformin, and thiazolidinediones.^30-40Summary and Conclusions
Considerable evidence has accumulated over the past 40 years indicating that glucagon, along with insulin, is a major regulator of glucose homeostasis in humans.

This important role derives from:

1. the need for glucagon to sustain appropriate glucose release by the liver in the fasting state in order to counterbalance the action of insulin and prevent hypoglycemia; and
2. the need for suppression of glucagon release after meal ingestion to permit the replenishment of hepatic glycogen stores through the action of insulin and the minimization of postprandial hyperglycemia.

It is now widely recognized that secretion of glucagon is increased in both type 1 and type 2 diabetes and that this imbalance between glucagon and insulin availability leads to greater hyperglycemia in both the fasting and postprandial states than can be explained by insulin deficiency alone. Agents such as GLP-1 analogs and DPP-4 inhibitors, which overcome reduced endogenous GLP-1 secretion found in type 2 diabetes, have beneficial effects on glycemic control, which at least in part can be attributed to their suppression of glucagon secretion.