| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Benaroya Research Institute, Seattle, Washington 98101
Address all correspondence and requests for reprints to: Dr. Carla J. Greenbaum, Benaroya Research Institute, 1201 9th Avenue, Seattle, Washington 98101. E-mail: cjgreen{at}benaroyaresearch.org.
As long as there have been tools to measure and manipulate glucagon accurately in experimental animal models and humans, there has been controversy as to the contribution of glucagon to fasting and postprandial hyperglycemia in the diabetic state. In healthy subjects, the actions of glucagon counteract and are tightly coordinated with insulin and other counterregulatory hormones to regulate blood glucose levels in a narrow range via regulation of hepatic glucose production through glycogenolysis or gluconeogenesis (1). There is agreement that in both type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM), glucagon is inappropriately elevated in the presence of hyperglycemia (2). In addition, particularly in T1DM, glucagon does not increase in the presence of hypoglycemia (3). In T1DM, the inappropriate elevation of glucagon during hyperglycemia has been attributed to the lack of intraislet insulin to restrain glucose’s effect on the 
-cell (4). In T2DM, in which insulin is present but unable to maintain normal glucose levels, impaired sensitivity of both the ß- and -cell to glucose has been proposed (5). An alternative concept, that the -cell is relatively insulin resistant in T2DM, is consistent with the proposed mechanism occurring in T1DM, specifically that intraislet insulin is unable to restrain the effect of glucose on the -cell.
Many experimental models have explored the relative roles of insulin, glucose, and glucagon. These include experiments in humans in which endogenous glucagon and insulin are suppressed with somatostatin and then selectively infused at various time points and concentrations to evaluate the impact on glucose (reviewed in Ref. 6). In general, these experiments provide evidence that glucagon, independently of insulin, contributes to both fasting and postprandial hyperglycemia, although there is controversy particularly around the latter (7). We evaluated this directly in a canine model in which alloxan-induced insulin deficiency was isolated to the ventral lobe of the pancreas, allowing the dog to remain chronically euglycemic. We administered a stepped glucose infusion to study the effects of acute increases of glucose, insulin, or both on glucagon secretion from the ß-cell-deficient lobe. These experiments indicated that insulin was required for the effect of glucose on the -cell but provided no information on the mechanism by which this occurred (8). Furthermore, it is clear from data from islet transplant recipients that local insulin is not the whole story because the glucagon response remains aberrant despite local secretion (9). It has been suggested that this is due to the lack of normal islet innervation, although other explanations including delivery of islets into the liver and disruption of islet architecture itself may be involved.
Other tools have been recently used to understand the contribution of glucagon to fasting and postprandial glucose values. The phenotype of lowered glucose values in glucagon receptor knockout (10, 11) and other transgenic animals (12, 13) as well as studies with antiglucagon antibodies that demonstrate reduction in hyperglycemia in animals with residual ß-cell function (14) all support at least some direct role of glucagon. Recently in vitro data of isolated, cultured -cells have suggested that glucose can directly result in glucagon secretion (15). These data raise questions regarding the direct contribution of the -cell to fasting and postprandial hyperglycemia in diabetes. Does glucose directly stimulate glucagon secretion and result in worsening hyperglycemia? Or is postprandial hyperglycemia due to the lack of insulin alone?
In this issue of JCEM, Porksen et al. (16) studied glucagon, glucagon-like peptide (GLP)-1, C-peptide, and glucose in response to a mixed meal in 257 children with T1DM at 1, 6, and 12 months after diagnosis with an aim to understand the role of glucagon in the development of postprandial hyperglycemia over time as individuals lost endogenous ß-cell function. They found that stimulated C-peptide falls about 50% from baseline during the first year after diagnosis and that postprandial glucose, glucagon, and GLP-1 increase during that period of time. Using a compound symmetrical repeated measurement model of all time points, the increasing levels of glucagon were found to be independent of age, gender, C-peptide, and fasting glucose yet highly associated with postprandial glucose. A separate model, evaluating these relationships at each time point studied indicated no significant relationship of glucagon and postprandial glucose soon after diagnosis (when ß-cell function is present), yet by 6 and 12 months marked postprandial hyperglycemia was associated with increased glucagon levels. Similar findings were found with GLP-1 and postprandial glucose. The authors suggest that these data support the concept that hyperglycemia directly results in glucagon secretion. Further support for this concept comes from immunohistochemical staining of human -cells, in which they find colocalization of Kir6.2 and SUR1, indicating that an ATP-sensitive potassium channel exists in both - and ß-cells.
Several caveats are important in interpretation of the data of Porksen et al. (16). First, as designed, the study describes only associations, not causality. The association of postprandial hyperglycemia and postprandial glucagon levels does not address the question as to whether the hyperglycemia resulted in elevation of glucagon or whether the elevated glucagon as a result of insulin deficiency augmented the hyperglycemia observed. Nonetheless, this study provides very important prospective data on changes in glucagon concentrations in subjects within the first year after diagnosis and implies that pronounced insulin deficiency is required before glucagon dysregulation is present. This conclusion is somewhat in contrast to data we obtained in a small number of subjects in which inappropriately elevated glucagon levels in response to oral or iv glucose were seen in individuals with very early T1DM (at a time in which fasting glucose values remained within the normal range and insulin secretion, although impaired, was still present) (17). Studies with more frequent data points may be required to understand better the true natural history of glucagon secretory responses in T1DM.
Perhaps more provocative is the notion that inhibition of the Kir6.2/Sur1 channel by sulfonylurea treatment might directly increase glucagon secretion and thus be potentially useful in the treatment of insulin-induced hypoglycemia in individuals lacking intraislet insulin secretion. Clearly, further study is needed before one considers this approach.
A clinician may wonder whether it is important to understand whether there is a glucose-sensing mechanism on the -cell and whether hyperglycemia directly increases glucagon secretion. In this context it is worthwhile to note that several therapeutic agents already in clinical use to treat diabetes are known to impact glucagon. In addition to delaying gastric emptying and acutely increasing insulin secretion, the GLP-1 mimetics (Exenatide, Amylin Pharmaceuticals, San Diego, CA; Sitagliptin, Merck, Whitehouse Station, NJ; Vildagliptin, Bristol-Myers Squibb, New York, NY) decrease postprandial glucagon concentrations. Their impact on glucagon has been suggested to contribute to their salutary effects in reducing postprandial hyperglycemia (reviewed in Ref. 18). Furthermore, therapeutic use of amylin (Byetta; Amylin Pharmaceuticals), a ß-cell hormone cosecreted with insulin and thus absent in individuals with long-standing T1DM, also results in delay of gastric emptying and reduction of both postprandial glucagon levels and hyperglycemia. Additional molecules that block glucagon secretion are in clinical development (19).
Notwithstanding these well-documented effects, it is of interest that the impact of these therapies on hemoglobin A1c (HbA1c) has been underwhelming. A reduction of HbA1c of less than 1% in type 2 subjects, particularly in light of the associated weight loss seen with Exenatide, suggests that preventing postprandial glucagon secretion may not have a dramatic clinical impact on individuals with diabetes (20). It is possible that a more pronounced clinical effect would be seen with specifically targeted therapies such as glucagon receptor antagonist peptides or small molecules. Such a view is also supported by the analysis performed by Porksen et al. (16) in which they failed to find a statistically significant association between postprandial glucagon levels and glycemic control. Unfortunately, these investigators devised an unconventional measure of glycemic control for this analysis; specifically, they calculated an insulin dose-adjusted HbA1c value derived from a regression analysis of their data with C-peptide as the dependent variable and then used the resulting formula of HbA1c + 4 x insulin dose as a measure of glycemic control tested in the same data set. Thus, it is not clear that this observation adds much to our understanding about the impact of postprandial glucagon secretion on glycemic control.
In summary, the data presented by Porksen et al. (16) are consistent with, but do not resolve the question whether, in the absence of sufficient insulin, marked hyperglycemia directly results in glucagon secretion that further contributes to hyperglycemia in the postprandial state. Nonetheless, their data provide important descriptive information about the natural history and relationships of glucagon, insulin, GLP-1, and C-peptide secretion in a large number of T1DM subjects tested in a standardized manner. More direct demonstration of the clinical impact of this hypothesis likely awaits development of therapies directly blocking the effects of glucagon.
Footnotes
Abbreviations: GLP, Glucagon-like peptide; HbA1c, hemoglobin A1c; T1DM, type 1 diabetes mellitus ; T2DM, type 2 diabetes mellitus.
Received June 12, 2007.
Accepted June 14, 2007.
References
-cell abnormality in human diabetes. Metabolism 25(11 Suppl 1):1527–1529-cell dysregulation in fasting and postprandial hyperglycemia in type 2 diabetes and therapeutic implications. Endocr Rev 28:253–283 cell hyperplasia in glucagon receptor knockout mice. Proc Natl Acad Sci USA 100:1438–1443-cell hyperplasia in prohormone convertase 2 knockout mice. Diabetes 51:398–405 cells. Diabetologia 50:370–379[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
