Hepato-Incretin Function of GLP-1: Novel Concept and Target in Type 1 Diabetes

The liver is a key player in glucose regulation, and several factors are involved in this regulation (1). Glucagon, autonomic nerves, and the adrenals stimulate hepatic glucose production to allow sufficient brain glucose delivery. Insulin inhibits hepatic glucose production, preventing excess hepatic glucose release. This is of particular importance after meals, as is evident from the negative correlation between early insulin response after meals and postprandial glucose (2). Therefore, low insulin and high glucagon levels—such as observed in type 1 and type 2 diabetes—sustain liver glucose production, thereby resulting in postprandial and fasting hyperglycemia. Thus, reducing hepatic glucose production is an important target in the treatment of diabetes (3).

The incretin hormone GLP-1 inhibits hepatic glucose production through its ability to stimulate insulin secretion and inhibit glucagon secretion (4). As GLP-1 is released after meal ingestion in proportion to meal size (5), this effect allows meal-dependent inhibition of hepatic glucose production. However, GLP-1 may also inhibit endogenous glucose production through an islet-independent mechanism (6–9). This allows not only a rapid shutoff of endogenous glucose production after meal ingestion but also incretin therapy to be potent in insulin-deficient diabetes.

A study in the current issue of Diabetes presents convincing evidence for such an islet-independent inhibition of endogenous glucose production by GLP-1 (10). The study by Jun et al. (10) leveraged the unique glucagon receptor mouse knockout model. These mice had low baseline glucose levels, favorable glucose tolerance, hyperglucagonemia, and α-cell hyperplasia (11). They were also resistant to induction of insulin-deficient diabetes (12). The new study showed that glucose levels were enhanced when the β-cells were chemically destroyed by streptozotocin in mice with genetic deletion of both glucagon and GLP-1 receptors (10). Furthermore, glucose levels after streptozotocin administration to GLP-1 receptor knockout mice were markedly increased by administration of a high-affinity glucagon receptor antibody. Thus, loss of GLP-1 action increases glucose levels and worsens glucose tolerance in insulin-deficient mice with no glucagon action. These results are similar to a recent finding that glucose levels are elevated by administering the GLP-1 receptor antagonist exendin_(9-39) to glucagon receptor knockout mice that have been rendered insulinopenic by streptozotocin (13). Jun et al. (10) proceeded further, however, by examining endogenous glucose production. This was increased in mice that lacked both glucagon and GLP-1 receptors after destruction of β-cells with streptozotocin. Taken together, the results from this interesting study support the idea that the resistance of glucagon receptor knockout mice to streptozotocin-induced diabetes is partially explained by GLP-1. This occurs through an insulin-independent action of this hormone that inhibits endogenous glucose production.

The strength of this article is the visionary and innovative use of a double knockout of both glucagon and GLP-1 receptors. An obvious limitation of the study is that it is undertaken in animals, highlighting the need for carefully designed human studies to explore the potential of such an important action of GLP-1 in humans.

The findings of this study have three exciting implications. First, direct suppression of endogenous glucose production by GLP-1 provides a mechanism for a rapid inhibition of endogenous glucose production after meal ingestion. The initial shutoff of liver glucose production after meals may therefore be not only a consequence of islet hormonal actions but also instituted by GLP-1 reaching the liver after meal ingestion. If so, this would ensure very tight regulation of hepatic glucose production after meals, such that meal size, through GLP-1, regulates inhibition of hepatic glucose production.

Second, the study provides a mechanism for a glucose-lowering action of incretin therapy in type 1 diabetes. GLP-1 receptor agonists, liraglutide and exenatide (14,15), and the dipeptidyl peptidase-4 inhibitor vildagliptin (16) have demonstrated glucose-reducing effects in type 1 diabetes, which have been thought to be mainly due to inhibition of glucagon levels. In fact, the first evidence of reduced glucagon levels by GLP-1 in type 1 diabetes was reported more than 20 years ago (17). However, the results of the current study suggest that inhibition of endogenous glucose production independent of glucagon may be an additional mechanism of GLP-1.

Third, and perhaps the most exciting implication of this study, incretin therapy may be combined with glucagon antagonism for a powerful glucose-lowering strategy in type 1 diabetes. This would allow the use of submaximal doses of the two strategies, resulting in maximal hepatic glucose production and avoidance of adverse events (3).

However, it is still important to explore the mechanism of the islet-independent action of GLP-1 to inhibit endogenous glucose production. The mechanism could be a direct effect on hepatocytes, although this is unlikely as there is no robust demonstration of GLP-1 receptors in these cells. It is therefore more likely that the effect is executed through the autonomic nerves. It has been shown that GLP-1 receptors are expressed both in the nodose ganglia and in nerve terminals innervating portal vein autonomic nerves (18). This may initiate neural responses, contributing to not only stimulated insulin secretion—the neuro-incretin effect (19)—but also inhibition of hepatic glucose production. Indeed, central activation of GLP-1 receptors inhibits endogenous glucose production through neural effects (20). Further analyses of neural contribution to hepatic effects of GLP-1 can be performed by examining denervated and autonomic blocking models in humans and animals as well as in animals with targeted neuronal GLP-1 receptor knockout. In humans, studies on effects of GLP-1 on hepatic glucose production during various portal insulin-to-glucagon ratios are also warranted.

In summary, this novel study by Jun et al. (10) shows important synergistic contributions of GLP-1 receptor agonism and glucagon receptor antagonism to prevent hyperglycemia in the setting of insulin deficiency. Based on the new data, it may be proposed that a hepato-incretin function of GLP-1 exists, as is illustrated in Fig. 1. Such a potential hepato-incretin function of GLP-1 is of great interest both for the understanding of the physiology of GLP-1 and the potential utility of incretin therapy in both type 1 and type 2 diabetes.