New concept of the glucagon‐like peptide‐1 signaling pathway on pancreatic insulin secretion

Abstract

The intracellular glucagon‐like peptide‐1 (GLP‐1) signaling pathway, which involves cyclic adenosine monophosphate (cAMP), exchange protein directly activated by cAMP, cAMP‐dependent protein kinase A (PKA) and adenosine triphosphate‐sensitive potassium channels, has been widely accepted as a common mechanism of GLP‐1‐stimulated insulin secretion. Recent studies showed that a stimulatory effect of GLP‐1 is also mediated by cAMP/PKA‐independent mechanisms, including induction of Gαq activity followed by phospholipase C and protein kinase C activation. Furthermore, transient receptor potential 4 and transient receptor potential 5 channels play a role in protein kinase C‐induced Ca²⁺ current. This pathway is a unique action mechanism of GLP‐1 at physiologically low concentrations.

Glucagon‐like peptide‐1 (GLP‐1), a member of the incretin family, is a hormone released from L cells in the distal intestine/colon after a meal. GLP‐1 analogs or its receptor agonists are widely used as antidiabetic drugs, which stimulate glucose‐dependent insulin secretion. The active form of GLP‐1 (7–36) is quickly degraded to an inactive form (9–36) by dipeptidyl peptidase‐4 (DPP‐4), mainly in the intestine. Thereby, a steady‐state physiological concentration of GLP‐1 is <10 pmol/L in the peripheral blood, and not >30 pmol/L even after a meal1. In fact, DPP‐4 inhibitors increase plasma concentrations of GLP‐1 only up to ≤20 pmol/L in diabetes patients, yet significantly enhance insulin secretion from the pancreas.

In contrast, many studies investigating the effects of GLP‐1 on pancreatic islet function have been carried out at nanomolar levels of GLP‐1, which are at least 100‐fold higher than those seen physiologically. Historically, such high concentrations of GLP‐1 have been used customarily in in vitro studies for the cellular mechanism of this hormone action. Actually, an EC₅₀ of a few nmol/L was suggested in the receptor‐binding assay and intracellular cyclic adenosine monophosphate (cAMP) accumulation. High concentrations (≥0.1 nmol/L) of GLP‐1 induce the receptor coupling to the guanosine 5′‐triphosphate‐binding protein Gαs, which activates adenyl cyclase and is followed by activation of the cAMP–dependent protein kinase A (PKA) pathway. The intracellular GLP‐1 signaling pathway, which involves cAMP, exchange protein directly activated by cAMP (Epac2), PKA and mediated channels, has been widely accepted to explain a mechanism of GLP‐1‐stimulated insulin secretion2.

As physiological levels (pmol/L) of GLP‐1 are much lower than those used in the studies (nmol/L) for the aforementioned GLP‐1 signaling pathway, the following question is raised: whether a cAMP–PKA‐dependent pathway shown at high concentrations of GLP is universal as an action mechanism of this hormone or not. Our previous in vitro study might provide a clue to deal with the question3. The study supported a possible effect of low concentrations (pmol/L) of GLP‐1, which was seen in the case of DPP‐4 inhibitor use, on insulin secretion. In that study, a high concentration (10 nmol/L) of GLP‐1 increased insulin secretion from MIN6 cells with a significant increase in intracellular cAMP accumulation. In this condition, TK5720, a selective PKA inhibitor, suppressed insulin secretion. In contrast, a low concentration (1 pmol/L) of GPL‐1 also stimulated insulin secretion without a significant accumulation of intracellular cAMP. Insulin secretion was not affected by a PKA inhibitor, but was inhibited by calcium channel blockers, such as verapamil and dantrolene, and an intracellular Ca²⁺ chelating agent, 1,2‐Bis(2‐amino‐phenoxy)‐ethane‐N,N,N’,N’‐tetraacetic acid. These data showed that picomolar GLP‐1‐induced insulin secretion is dependent of intracellular calcium concentration, but independent of the cAMP–PKA pathway. In response to these results, we speculated on a new possibility for a unique mechanism of picomolar GLP‐1 action different from that shown at high nanomolar concentrations4.

The recent study reported by Shigeto et al.5 has made the possibility real. First of all, the study examined the dose‐dependent properties of GLP‐1 effects on glucose‐stimulated insulin secretion (GSIS) in mouse and human islets. It was again confirmed that GLP‐1 between 0.1 pmol/L and 10 pmol/L enhanced GSIS dose‐dependently with EC₅₀ of ~0.4 pmol/L in isolated mouse islets. A potency of 1 pmol/L GLP‐1 was almost the same as that of high concentrations (~10 nmol/L). In experiments to assess electrical activity in β‐cells, the membrane potential generated in dispersed β‐cells exposed to physiological concentration (6 mmol/L) of glucose was observed in just 10% of the total number of cells, but the addition of 1 pmol/L GLP‐1 significantly and reversely initiated or enhanced action potential firing in almost whole cells. The potency of 1 pmol/L GLP‐1 on the electrical activity in β‐cells was almost equal to that evoked by a higher concentration (10 nmol/L) of this hormone.

In the presence of 6 mmol/L glucose, the induction of [Ca²⁺]i oscillations was observed in just 5% of the cells, which was similar to the fraction of cells shown in the effect on electrical activity. However, 1 pmol/L GLP‐1 again significantly increased the recruitment of cells, in which a spontaneous Ca²⁺ current was induced. Exocytosis of insulin granules evoked by the membrane depolarization was also enhanced by the addition of 1 pmol/L GLP‐1. As these effects of 1 pmol/L GLP‐1 were inhibited by the GLP‐1 receptor antagonist, exendin (9–39), it was strongly suggested that the stimulatory effect of physiologically low levels of GLP‐1 on GSIS would be mediated by the GLP‐1 receptors that are currently recognized.

Shigeto et al.5 further showed that 1 pmol/L GLP‐1 potentiated the integrated Ca²⁺ current in mouse β‐cells. This effect was thought to be mediated by the activation of L‐type Ca²⁺ channels, because isoradipine, an inhibitor of these channels, significantly reduced the Ca²⁺ current induced by 1 pmol/L GLP‐1. GSIS potentiated by a low concentration of GLP‐1 was also inhibited by isoradipine. These results were confirmed in human β‐cells. In the experiment using the membrane permeable PKA inhibitor, myristoylated PKI or 8‐Br‐Rp‐cAMP, the stimulatory effect of 1 pmol/L GLP‐1 on GSIS was significantly inhibited, indicating the possibility that the stimulatory effects of picomolar GLP‐1 on insulin secretion are mediated by a PKA‐independent mechanism. In contrast, 1 pmol/L GLP‐1 stimulated cAMP accumulation in renal COS7 cells transfected with human GLP‐1 receptor, suggesting that the cAMP–PKA‐dependent mechanism also remained to a certain degree.

From the above process, a PKA‐independent/PKC‐dependent mechanism was found to be involved in the stimulatory effects of picomolar GLP‐1 on insulin secretion. Then, it has been suggested that GLP‐1 receptor binding activates not only Gαs, but also Gαi and Gαq, which are linked with phospholipase C (PLC) and protein kinase C (PKC) activation. The imaging of cytosolic/submembrane diacylglycerol (DAG), a PKC activator, strongly supported an involvement of PLC/PKC, which played a significant role in picomolar levels of GLP‐1‐stimulated insulin secretion. Furthermore, a physiological (pmol/L) GLP‐1‐induced adenosine triphosphate‐sensitive‐independent depolarization was closely related with an activation of a Na⁺‐permeable conductance, using an electrophysiological technique5.

Suggested Na⁺‐permeable conductance was activated by a low level of GLP‐1, and transient receptor potential (TRP) channels were assumed to mediate GLP‐1 effects on insulin secretion. TRP cation channel subfamily M member 4 (TRPM4) and TRPM5 were chosen from several TRPMs expressed in pancreatic β‐cells. These two TRPM, which are Ca²⁺‐sensitive cation‐conducting channels and share approximately 40% amino acid identity, are functionally similar to each other. TRPM4 is known to be involved in the PKC pathway. Interestingly, the picomolar GLP‐1‐induced membrane depolarization was significantly suppressed in Trpm4−/− or Trpm5−/− mice⁵.

A line of evidence draws the conclusion that a physiologically low level (pmol/L) of GLP‐1, which is enough to stimulate insulin secretion from β‐cells, affects cell membrane potential by collaboration of Na⁺ current activation through TRPM4 and TRPM5 channels involved in the PKC‐dependent pathway and PKA‐dependent reduction of adenosine triphosphate‐sensitive channel activity. Thus, it is clearly shown that the effects of picomolar GLP‐1 on insulin secretion are mediated by the PKC‐dependent pathway, in addition to a PKA‐dependent mechanism as a common pathway of GLP‐1 action.

Here, I advocate a new concept of the intracellular GLP‐1 signaling pathway on insulin secretion (Figure 1). Low (~30 pmol/L) concentrations of GLP‐1 stimulate insulin secretion from pancreatic β‐cells in the presence of physiological levels of glucose accompanied by membrane depolarization and an increase in the intracellular [Ca²⁺]i via L‐type Ca²⁺ channel. Unlike nanomolar levels of GLP‐1, this response is mediated in a considerable part by the PKA‐independent pathway, including induction of Gαq activity followed by PLC and PKC activation. Furthermore, TRPM4 and TRPM5 channels play a role in the PKC‐induced Ca²⁺ current. This pathway is a unique action mechanism of GLP‐1 at physiologically low concentrations.

Figure 1.

A new concept for intracellular signaling pathway of glucagon‐like peptide‐1 (GLP‐1)‐stimulated insulin secretion. A common pathway of GLP‐1 in any concentration is drawn in black and a unique pathway of GLP‐1 at physiologically low levels (pmol/L) is in red. AC, adenyl cyclase; DAG, diacylglycerol; IP3, inositol triphosphate; PKA, protein kinase A; PKC, protein kinase C; PLC, phospholipase C; TRPM, transient receptor potential cation channel subfamily M member.

The reason why an intracellular mechanism of GLP‐1 action is dependent of its concentration is not easy to explain clearly. GLP‐1 is known to work in many ways as a hormone in the pancreas and as neuropeptides in the brain. Although the actual concentrations of GLP‐1 in each organ are not always definite, it is speculated that GLP‐1 functions at suitable concentrations according to the purpose in its target organs.

The concept that some receptor ligands act through distinct signaling pathways is well known; for example, epinephrine binding to its α‐receptor activates not only Gαs, but also Gαi according to the role.

Furthermore, it has not been excluded that GLP‐1 stimulates insulin secretion neurally; that is, through activation of vagovagal reflexes, suggesting its concentration is higher than that seen in peripheral blood. However, a series of events clearly showed that picomolar levels of GLP‐1 are physiologically relevant and sufficient to stimulate insulin secretion, even if the contribution of the PLC–PKC pathway in picomolar GLP‐1 action might be different in each animal type. Each different intracellular signaling pathway observed at high and low concentrations of GLP‐1 presents a hint to understanding a multifunctional property of this hormone in various organs.

Disclosure

The author has received advisory fees from Astellas Pharma, Sanwa Kagaku Kenkyusho, Novo Nordisk Pharma and Takeda Pharmaceutical; honoraria from Astellas Pharma, AstraZeneca, Daiichi Sankyo, MSD, Ono Pharmaceutical, Novo Nordisk Pharma, Boehringer Ingelheim Japan, Taisho Toyama Pharmaceutical, Takeda Pharmaceutical and Mitsubishi Tanabe Pharma; and scholarship/donation fees from Boehringer Ingelheim Japan, Taisho Toyama Pharmaceutical, and Mitsubishi Tanabe Pharma and Kowa Pharmaceutical.