Brief overview: glucagon history and physiology

Introduction

The editors asked that I write a brief introductory article for this special collection, which focuses on scientific advances in glucagon research and its clinical impacts. In doing so, I will provide a short history of glucagon discovery and research, and then briefly describe several examples of its far-reaching physiology in the context of its interactions with its partner hormone insulin. But to be certain, the material you hold in your lap or gaze at on your screen is principally about glucagon. In my writing, I reference the Broadway play entitled ‘The Odd Couple’ for several reasons. The play was written in 1965 by Neil Simon about two men living in different bedrooms in the same apartment and who were in most ways the opposite of each other. So, too, with glucagon and insulin, which are synthesized in different cells in the same compartment (the islet), as you will read. There is also a coincidental aspect with this metaphor because the two main actors in The Odd Couple were named Unger (Felix) and Madison (Oscar).

Brief history

Glucagon is a peptide hormone that is produced primarily by the alpha cells in the islet of Langerhans in the pancreas, but also in the stomach, intestinal enteroendocrine cells, and some neurons. Approximately 100 years ago, several research groups (Muller et al. 2017) discovered that pancreatic extracts would cause a brief rise in blood glucose before they observed the decrease in glucose attributed to insulin. Kimball and Murlin identified this substance in the extracts and named it ‘glucagon’, derived through the words ‘glucose agonist’ (Kimball & Murlin 1923, Muller et al. 2017). Eli Lilly is credited with the isolation of pure glucagon and the determination of its amino acid sequence (Kimball & Murlin 1923, Staub et al. 1953 Bromer et al. 1957, Muller et al. 2017). The precursor of glucagon is proglucagon, which is encoded by the GCG (glucagon) gene. Proglucagon is cleaved by two prohormone convertases. Prohormone convertase-2 is expressed in the alpha cell and the major product of proglucagon cleavage is glucagon (Orskov et al. 1987). On the other hand, prohormone convertase-1/3 is mainly expressed in the intestinal enteroendocrine cells and cleaves proglucagon to form oxyntomodulin, glucagon-like peptide-1 (GLP-1) and GLP-2. The research group led by Roger Unger, a member of which was the master clinician Leonard Madison, developed the first radioimmunoassay for measuring blood glucagon levels in 1959 (Unger & Orci 1981). Unger’s group became the major leader in the development of glucagon physiology and pathophysiology research throughout the ensuing 50 years.

Brief overview of the physiology and regulatory effects of glucagon

Returning to the Odd Couple theme, an overall description of the regulation of glucagon secretion requires consideration of its neighbor insulin because they both are made primarily by the islet and they both are capable of affecting the function of each other in different ways. For example, in vivo glucagon can stimulate insulin secretion, whereas insulin and glucose suppress glucagon secretion. The mechanism of action of glucagon on insulin secretion has been identified as a trimeric guanine nucleotide-binding protein (G-protein)-mediated event (Wu et al. 2021). Glucagon bound to its receptor interacts with the alpha-subunit of a trimeric G-protein on the cytoplasmic surface of the beta cell. The G-protein complex exchanges GDP for GTP, which activates the alpha subunit, which in turn stimulates insulin secretion. This event may be thought of as primarily pharmacologic in nature, but possible physiologic relevance has not been ruled out.

Role of insulin in suppression of glucagon release

The mechanism whereby insulin suppresses glucagon release from the alpha cell is thought to be highly dependent on the peri-portal circulation of the islet through which blood flows predominantly downstream from beta cells to alpha cells, although this classical model for the direction of islet blood flow has recently been challenged. In this scenario, it is via the circulation that insulin is thought to suppress the release of glucagon. Another consideration is the paracrine effect of insulin independent of blood flow. Additionally, high levels of glucose in vivo also have been shown to suppress glucagon secretion. Consequently, the glucose-lowering effect of insulin may be permissive or additive to the effects of glucose to suppress alpha cell function so that the distinction between the lowering of glucagon levels by insulin and glucose in intact mammals, including humans, is complex. This issue was addressed by Asplin et al. (Asplin et al. 1981) who compared the acute glucagon response to intravenous arginine in the basal state and after beta cell suppression by intravenous infusion of insulin using a glucose clamp. This method of endogenous beta cell suppression was associated with increased glucagon responses to arginine. Control experiments using C-peptide-negative subjects with type 1 diabetes, who presumably had no endogenous insulin secretion, had no such augmentation of arginine-induced glucagon secretion. The authors concluded that inhibition of beta cell function and insulin secretion potentiates glucagon secretion via a paracrine mechanism. In other studies, Greenbaum et al. tested whether insulin is permissive for the direct suppression of alpha cell function by glucose, and/or whether secretion of the delta cell-derived inhibitory hormone somatostatin is required (Greenbaum et al. 1991). These investigators compared inhibition of glucagon secretion during a stepped intravenous glucose infusion in dogs that had ventral lobe-specific beta cell deficiency, achieved by alloxan treatment of the vascularly clamped ventral pancreas. The rest of the pancreas was not alloxan-treated and received normal blood flow. In control dogs, glucagon suppression occurred at a glucose level of 150 mg/dL, with no effect on somatostatin; somatostatin output only increased at glucose levels > 250 mg/dL. In alloxan-treated dogs, glucagon output from the clamped ventral lobe was not suppressed nor was somatostatin output increased in response to the stepped intravenous glucose infusion. However, when insulin and glucose were infused directly into the pancreatic artery that supplied the alloxan-treated lobe, they restored the ability of hyperglycemia to suppress glucagon and stimulate somatostatin secretion. This suppression was not seen when insulin alone without glucose infusion was infused. Overall, the authors concluded that intra-islet insulin is permissive for glucose to suppress alpha cell function.

In regard to the above experiments, isolated islets, unlike islets in intact pancreata in vivo, have repeatedly been shown not to release glucagon when they are exposed to very low concentrations of glucose in vitro. This has been variously attributed to lack of innervation or damage to alpha cells during isolation procedures. Based on studies by others (Samols et al. 1972, Banarer et al. 2002) of intraislet regulation of glucagon secretion by insulin, we proposed that one might identify an ‘insulin switch-off’ phenomenon (Hope et al. 2004, Zhou et al. 2004) by using perifusion experiments. We posited that the lack of periportal blood flow in isolated islet preparations renders the delivery of insulin to downstream alpha cells, and in turn provision of an insulin switch-off signal, impossible. As expected, exposure of normal rats or human islets to low glucose levels during islet perifusion with saline alone or with low glucose concentrations did not result in glucagon release. However, when perifused islets were first exposed to a high glucose concentration in the perfusate to stimulate insulin secretion, and then followed with stopping the glucose infusion to provide an insulin switch-off signal, we observed a glucagon response. This response was not observed if the insulin perifusion was not switched off when the islets were deprived of glucose or when insulin was switched off during glucose deprivation. We concluded that both the condition of switching off the surrogate intravascular insulin signal and the condition of low glucose were responsible for the induction of glucagon secretion.

Metabolic effects of glucagon

In addition to the complex interrelationships involving glucagon, insulin, and levels of glycemia, in the past 50 years, many investigators have reported physiologic mechanisms through which glucagon influences, and is influenced by, other metabolic events (Table 1). For example, it is well known that epinephrine release during stress states stimulates glucagon secretion by virtue of its beta-adrenergic property and inhibits insulin by virtue of its alpha-adrenergic property (Robertson & Porte 1973). Increased glucagon secretion and increased glucose production during states of systemic stress, with changes in alanine kinetics, have been observed. In terms of stress, Jahoor et al. examined the role of insulin and glucagon in the response of glucose and alanine kinetics in burn-injured patients by infusing somatostatin with and without insulin (Jahoor et al. 1986). They observed that selectively lowering glucagon concentration with somatostatin lowered glucose production while the addition of an exogenous glucose infusion was successful in maintaining euglycemia. In this context, amino acids are known to stimulate both insulin and glucagon secretion and studies in this area suggest that glucagon and amino acids may be involved in a mutual feedback cycle (Holst et al. 2017). Other studies have examined the effects of glucagon on lipolysis and ketogenesis. Liljenquist et al. examined the effects of glucagon on arterial glycerol concentration and net splanchnic production of total ketones and glucose (Liljenquist et al. 1974). In normal subjects, glucagon stimulated C-peptide secretion, whereas arterial glycerol concentration and net splanchnic total ketones were markedly decreased. In contrast, while glucagon had no effect on C-peptide levels in insulin-dependent diabetic men, arterial glycerol and net splanchnic total ketone production rose significantly. They concluded that in intact man, glucagon, mole for mole, has more activity in stimulating hepatic glucose release than insulin has in opposing this process and that insulin on a molar basis has more antilipolytic activity than glucagon has lipolytic activity.

Therapeutic implications, with a focus on transplantation

In the 100 years since the discovery of glucagon and its subsequent investigation at the basic and clinical levels, it has achieved billing equal to that of insulin in terms of physiologic importance to a large variety of metabolic processes. These regulatory effects range from control of glycemia, lipolysis, ketogenesis, and amino acid turnover. There are also implications for clinical treatment of diabetes mellitus.

The widespread physiological roles of glucagon on glucose, lipid, and amino acid metabolism have led to the consideration of glucagon to play a role in the treatment of type 2 diabetes mellitus. For a single, recently published review describing this potential, the reader is referred to the 2012 publication by Unger and Cherrington (Unger & Cherrington 2012) that nicely encapsulates the Unger thesis that ‘glucose-responsive beta cells normally regulate juxtaposed alpha cells and that without intraislet insulin, unregulated alpha cells hypersecrete glucagon, which causes the symptoms of diabetes.’ This line of thinking has led to the question of whether drugs designed to inhibit glucagon action might be effective in the clinical management of diabetes patients. One such study was published in 2016 by Kazda et al. (Kazda et al. 2016) that evaluated this idea through a clinical trial of a small-molecule glucagon receptor antagonist in patients with type 2 diabetes. The drug was successful in modestly reducing levels of HbA1c and fasting glucose after 12 and 24 weeks of treatment.

In terms of type 1 diabetes, the therapeutic implications are of a different nature. In this case, it is not so much how glucagon administration might affect treatment as it is how treatment affects glucagon secretion during hypoglycemia. Pancreatic and islet transplantation are established methods of restoring insulin secretion in both type 1 and type 2 diabetes, although they are used much more frequently in type 1 diabetes. In both instances, this therapeutic approach is reserved for patients who have severe secondary complications of the disease that cannot be successively alleviated by more conventional therapeutic modalities. While successful transplantation will restore insulin secretory responses that are quantitatively normal when corrected for the amount of islet mass that is transplanted (Robertson et al. 2015), there can be important consequences for glucagon secretion during hypoglycemia. Pancreatic organ recipients achieve completely normal glucagon secretion from the transplanted pancreas during hypoglycemia (Diem et al. 1990). In this instance, the glucagon response is by the transplanted pancreas not the native pancreas. This is somewhat surprising because the vagal nerve is known to play a role in glucagon responses, and yet there are no meaningful vagal inputs to a transplanted pancreas, which is placed in the pelvic cavity. The other aspect that is notable is in the case of islet transplantation. Isolated islets are used for this procedure in which the islets are infused into the portal vein which delivers them to the liver sinusoids. Here, they have normal glucagon responses to amino acids but only partial responses to hypoglycemia (Paty et al. 2002, Rickels et al. 2015). In the case of autoislet transplantation, the recipient’s own islets are used after a complete pancreatectomy as a treatment for chronic, painful pancreatitis. These patients exhibit abnormal glucagon secretion during hypoglycemic clamps (Bellin et al. 2014). However, if the islets are placed in non-hepatic tissue, such as the omentum, the glucagon response is normal. The mechanism for this has been suggested to be the intimate relationship that intrahepatic alpha cells have with hepatic glycogen stores and release of free glucose due to the adrenergic stimulation by circulating epinephrine during hypoglycemia (Zhou et al. 2008) and/or adverse effects of a mismatch between insulin and glucose levels secondary to the gastrointestinal surgery that accompanies complete pancreatectomy. These issues surrounding the use of the liver as a transplantation site have important implications for fields developing other more experimental approaches to the restoration of islet function that propose using stem cell and xenogenic animal islet transplantation.

Conclusion

In the 100 years since the discovery of glucagon and its subsequent investigation at the basic and clinical levels, it has achieved billing equal to that of insulin in terms of physiologic importance to a large variety of metabolic processes. These regulatory effects range from control of glycemia, lipolysis, ketogenesis, and amino acid turnover. There are also implications for the clinical treatment of diabetes mellitus, including opportunities for pharmacological treatment of type 2 diabetes and considerations regarding the impact of pancreas and islet transplantation on glucagon release.

Declaration of interest

The author has no conflicts of interest to disclose.

Funding

This work was funded by NIH RO1 38325-35.