GLP-1 and GIP receptor signaling in beta cells – A review of receptor interactions and co-stimulation

Glucagon-like peptide 1 receptor (GLP-1R) and glucose-dependent insulinotropic polypeptide receptor (GIPR) are two class B1 G protein-coupled receptors, which are stimulated by the gastrointestinal hormones GLP-1 and GIP, respectively. In the pancreatic beta cells, activation of both receptors lead to increased cyclic adenosine monophosphate (cAMP) and glucose-dependent insulin secretion. Marketed GLP-1R agonists such as dulaglutide, liraglutide, exenatide and semaglutide constitute an expanding drug class with beneficial effects for persons suffering from type 2 diabetes and/or obesity. In recent years another drug class, the GLP-1R-GIPR co-agonists, has emerged. Especially the peptide-based, co-agonist tirzepatide is a promising candidate for a better treatment of type 2 diabetes by improving glycemic control and weight reduction. The mechanism of action for tirzepatide include biased signaling of the GLP-1R as well as potent GIPR signaling. Since the implications of co-targeting these closely related receptors concomitantly are challenging to study in vivo, the pharmacodynamic mechanisms and downstream signaling pathways of the GLP-1R-GIPR co-agonists in general, are not fully elucidated. In this review, we present the individual signaling pathways for GLP-1R and GIPR in the pancreatic beta cell with a focus on the shared signaling pathways of the two receptors and interpret the implications of GLP-1R-GIPR co-activation in the light of recent co-activating therapeutic compounds.


Introduction
Insulin is an essential hormone that facilitates glucose uptake in peripheral tissues, promoting blood glucose control, and basic cellular functions. Following nutrient ingestion, the postprandial insulin secretion is mainly caused by a rise in plasma glucose and a nutrient-induced secretion of the gastrointestinal hormones, glucagon like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) [1,2]. GLP-1 and GIP are responsible for the incretin effect, which is the potentiation of glucose-induced insulin secretion seen when nutrients pass the gastrointestinal tract (compared to intravenous nutrient administration) [3,4]. GLP-1 and GIP exert their physiological actions via stimulation of the two G protein-coupled receptors (GPCRs): the GLP-1 receptor (GLP-1R) and the GIP receptor (GIPR), respectively.

GPCRs of the human pancreatic beta cells
GPCRs are transmembrane proteins with seven alpha-helical segments separated by alternating intracellular and extracellular loop regions. The GPCRs are allocated in five families where the GLP-1R and GIPR are found within the secretin family, also classified as class B [31,32]. Upon stimulation by an extracellular stimuli (ligand), GPCRs undergo conformational changes, and triggers downstream intracellular signals by coupling with G proteins (or other intracellular proteins such as arrestins), causing a wide range of both physiological and pathological processes [33][34][35]. A ligand can activate or block an agonist-induced GPCR resulting in agonism or antagonism, respectively, of the receptor signaling. A ligand can also be biased towards one signaling pathway over another relative to a reference ligand. This means that receptor activation induced by a biased ligand e.g. can result in selective G protein recruitment and reduced arrestin recruitment compared to the endogenous ligand causing a diverse cellular response [36]. This concept has been shown for several agonists targeting the GLP-1R, such as the Val8-GLP-1 [37] and several variants of exendin 4 [11,38,39]. The GPCRs are highly attractive drug targets for many human diseases [40], and drugs targeting the GLP-1R and GIPR are in clinical development or already marketed as treatments of type 2 diabetes or obesity ( Table 1).
The beta cell is one of four cell types present in the islets of Langerhans, which are distributed throughout the endocrine pancreas. In response to hormones such as GLP-1 and GIP, nutrients, and neuronal stimuli the beta cells secrete insulin to maintain the plasma glucose levels in a small physiological range for optimal function of all tissues in the body [41,42]. The pancreatic beta cell is central for the physiological roles of GLP-1 and GIP and, thus, GLP-1R and GIPR signaling pathways. Glucose is the trigger of insulin secretion in beta cells, and GLP-1 and GIP both potentiate the glucose-stimulated insulin secretion via an increased cAMP concentration intracellularly resulting in an amplified insulin secretion [43]. Thus, the two agonists have small or no insulinotropic effect when blood glucose reach fasting concentrations, while they strongly potentiates insulin secretion during elevated (e.g. postprandial) blood glucose concentrations [44][45][46].

GLP-1 receptor signaling in the human beta cell
To stimulate insulin secretion, and in the presence of elevated blood glucose concentrations, GLP-1R activation in pancreatic beta cells promote recruitment and activation of G αs protein leading to adenylate cyclase-mediated cAMP production, elevation of Ca 2+ , and ERK1/2 phosphorylation (Fig. 3) [6,62,[76][77][78]. cAMP production lead to activation of protein kinase A (PKA) as well as exchange protein directly activated by cAMP (EPAC) and is directly involved in increasing proinsulin gene transcription and subsequent insulin secretion [79][80][81]. Activated PKA and EPAC enhance insulin granular exocytosis by insulin granular priming and phosphorylates sulfonylurea receptor (SUR1) K ATP channel subunit and thereby closes K ATP -channels in the plasma membrane [13,43,82], leading to membrane depolarization, and opening of the voltage gated Ca 2+ -channels. This leads to an increased influx of extracellular Ca 2+ , which triggers fusion of intracellular insulin-containing granules with the plasma membrane and thereby insulin secretion. Additionally, PKA activates transducer of regulated CREB (TORC2) and cAMP response element-binding protein (CREB), which leads to beta cell proliferation through CREB-and TORC2-mediated IRS2 gene expression [83,84]. In both rodent and human beta cells, EPAC2 also inhibits the function of the K ATP -channels through activation of SUR1 [83,85]. Furthermore, CRE activation leads to stimulation of beta cell proliferation through CREB-mediated IRS-2 gene expression [80,86]. The activated CREB also results in higher Bcl-2 activity, and inhibition of the pro-apoptotic bax, which is involved in beta cell survival processes and thereby maintain and promote beta cell functions [9,80]. GLP-1 also stimulate expression of GLUT2 transporters and glucokinase leading to an increased intracellular glucose Table 1 Established and future (potential) therapeutic compounds and widely used GLP-1R and GIPR ligands. Potencies and efficacies of ligands activating the glucagon-like peptide 1 receptor (GLP-1R) and/or glucose-dependent insulinotropic polypeptide receptor (GIPR) (in vitro studies). See references in the  [62] concentration and increased glycolysis [87].
For the non-G protein-dependent pathways, GLP-1R activation leads to c-SRC kinase-mediated transactivation of the epidermal growth factor tyrosine kinase receptor (EGFR) [13,84,87,88] (Fig. 3). This leads to activation of PI3K, which activates AKT/PK, leading to phosphorylation of the nuclear transcription factor Foxo1 [9,89]. Phosphorylated Foxo1 can inactivate PDX-1, which in murine beta cells, has been demonstrated to lead to anti-apoptotic effects [9,13,90,91]. Moreover, beta arrestin 1 binds to the GLP-1R upon activation by GLP-1 where the conformational receptor changes initiate phosphorylation of transmembrane helix 7 by G protein-coupled receptor kinases (GRKs) resulting in recruitment of beta arrestins. The beta arrestin recruitment results in ERK1/2 activation, beta cell proliferative and insulin secretion [6,80,92]. Additionally, beta arrestin 1 also plays a role in receptor desensitization by at least two mechanisms: binding to the receptor and, thus, preventing G protein recruitment and by promoting receptor trafficking by internalization and recycling [14,[92][93][94] (Fig. 3), although arrestin-independent internalization of the GLP-1R has been reported [95,96]. Furthermore, GLP-1R internalization is mediated by agonist-induced stimulation of the G αq pathway, which leads to receptor recycling to the plasma membrane or transport to lysosomes resulting in proteolysis. Beside this, G αq coupling leads to hydrolysis of PIP 2 via phospholipase C activation, an increased Ca 2+ accumulation, and phosphorylation of ERK 1/2 and beta cell proliferative processes [96]. Thus, multiple pathways are activated in the beta cell down stream of the GLP-1R.

GIP receptor signaling in the human beta cells
To stimulate insulin secretion after nutrient ingestion [16], GIP binds to and activates the GIPR on the beta cell surface. As for the GLP-1R, the activation of the GIPR promotes recruitment and activation of the G αs protein, which in turn activates the adenylate cyclase to stimulate cAMP production (Fig. 4). Again, the increased cAMP [97] activates PKA and EPAC [98]. PKA and cAMP activate a series of proteins including the mitogen-activated protein kinase (MAPK) cascades, and phosphorylates ERK1/2, which regulates genes involved in proliferative and anti-apoptotic processes [99,100]. The GIPR activation of PKA leads to insulin secretion by the same mechanisms as described for the GLP-1R [13,101,102]. GIPR activation can also promote non-insulinotropic actions, such as controlling pancreatic beta cell proliferation and survival (Fig. 4) [100,103]. Furthermore, PKA inhibits AMPK, which results in dephosphorylation and nuclear import of TORC2 [104]. CREB and TORC2 form a complex promoting the transcription of the anti-apoptotic gene bcl2 [104]. Activation of Akt/PKB/PI3K promotes phosphorylation of the nuclear transcription factor Foxo1, which in turn inactivates bax and other apoptosis-related factors, resulting in downregulation of the pro-apoptotic signaling pathway [105]. The upregulation of bcl2 and the   downregulation of bax leads to increased beta cell survival. Additionally, activation of Akt/PKB will result in suppressing the anti-proliferative and anti-survival mechanism of p38 MAPK and c-Jun N-terminal kinase (JNK), leading to suppressed mitochondrial translocation of BAD and BimEL and thereby the subsequent activation of caspase-3 promoting proliferation [100,103,104,106]. An important difference between GIPR and GLP-1R is that the GLP-1R activates EGFR leading to proliferation and anti-apoptosis, which is not the case for the GIPR [13].
When GIP is bound to its receptor, the receptor undergoes conformational changes that also initiate the phosphorylation of helix 7 by G protein-coupled receptor kinases (GRK). This leads to the recruitment of beta-arrestin 1 and 2 (Fig. 4). The beta-arrestins play a key role in receptor desensitization by blocking the G αs proteins and in receptor trafficking by internalization and recycling [14,62,107]. In contrast to the GLP-1R, where the internalization is both beta arrestin dependent and independent (e.g. G αq mediated pathway), the arrestins are necessary for GIPR internalization [107]. The coated pits that contain the GIPRs are separated from the plasma membrane by the membrane-remodeling GTPase dynamin leading to the construction of early endosomes, in which the GIPR is suggested to trigger production of cAMP and PKA activation [108] before recycling back to the beta cell surface.

Shared signaling pathways of the GIP and GLP-1 receptors in beta cells
Sharing the essential physiological task of stimulating postprandial insulin secretion, GLP-1R and GIPR in beta cells have many common intracellular pathways (Fig. 5A). Both receptors signal through activation of adenylyl cyclase resulting in increased intracellular cAMP, activating PKA and EPAC2 pathways which result in increased secretion of insulin, beta-cell proliferation, and survival/anti-apoptosis. Both GPCRs are desensitized by recruitment of beta-arrestins, followed by receptor internalization (though the GLP-1R internalization does not rely on on arrestins for internalization), recycling, and blockage of the G αs recruitment.
However, as for the differences, the activation of the EGFR pathway, leading to activation of PI3K and increased beta cell survival, is only activated by the GLP-1R and not the GIPR. Furthermore, the GLP-1R can be internalized by recruitment of beta arrestins and/or G αq [96] (Fig. 3) and the internalization process can be affected by several ligands [38,94] where the internalization process of the GIPR is not that easily affected by novel ligands but completely rely on arrestins [61,74]. Moreover, activation of the GIPR in beta cells has been shown to result in MAPK-induced signaling pathways, which is not seen for in the GLP-1R.

The rationale for GLP-1R-GIPR co-agonists
Due to the therapeutic potential of increased insulin secretion and improved beta cell health in patients with type 2 diabetes, GLP-1R-GIPR co-agonists are now in drug development programs [110]. Impressively, synergistic effects have been reported for a GLP-1R-GIPR co-agonist, tirzepatide, promoting higher insulin responses than separate administration of each hormone. Tirzepatide has potent glucose lowering and weight loss effects compared to the established GLP-1R agonists and demonstrates dose-dependent reductions in HbA1c in patients with type 2 diabetes [25,61,[111][112][113][114][115]. At the pharmacodynamics level, tirzepatide has been identified as an imbalanced GLP-1R-GIPR co-agonist implying a favor for GIPR over GLP-1R with equal affinity for the GIPR compared to endogenous GIP and slightly lower affinity for the GLP-1R than endogenous GLP-1. Also implied in the imbalanced profile of tirzepatide is a favor of cAMP generation over beta-arrestin recruitment at the GLP-1R which leads to weaker ability to drive receptor internalization compared with endogenous GLP-1 (Fig. 5B) [111]. This bias, relative to endogenous GLP-1, results in an increased amount of GLP-1R on the cell surface thereby increasing the intracellular signaling induced by the drug and stronger insulinotropic properties [11]. Based on in vitro receptor signaling studies, these tizepatide actions happen with concentrations of ~1− 10 nM reaching the maximal activity of up to 100 nM [111] which is the same concentration ranges as the native ligands (GLP-1 and GIP, respectively). Other GLP-1R-GIPR co-agonists are in preclinical/clinical development (Table 1) [60,61,73], and compared to tirzepatide, the GLP-1R-GIPR co-agonist HISHS-2001 has slightly higher GLP-1R and GIPR potencies (cAMP production) and lower beta arrestin-2 recruitment to the GLP-1R [60]. However, it is not yet known if the biased mode of action seen for tirzepatide and HISHS-2001, will be a defining drug class effect.

Receptor expression relationship
The therapeutic implications of GLP-1R-GIPR co-agonism not only rely of the impact of the individual receptor activation, but undoubtedly also on the interplay between the two receptors and their signaling pathways. As an example, the signaling profiles of cells co-expressing GLP-1R and GIPR (in a fixed receptor expression ratio) suggest that GIPR negatively impact GLP-1R expression and GLP-1R agonist actions [74]. In the same system, GLP-1-GIPR co-agonists primarily interfere with the beta arrestin recruitment of the GLP-1R and the G αs recruitment of the GIPR, inducing biased signaling as well as affecting the cell surface expressions. Moreover, co-expressed GIPR also severely reduce GLP-1R internalization, beta arrestin recruitment, and cAMP accumulation [116]. With this in mind, the classical pharmacodynamics term of agonism and antagonism, might not be sufficient to describe and determine the GLP-1R-GIPR targeting drugs needed e.g. to improve insulin secretion and beta cell function. For the GLP-1R-GIPR co-agonists tirzepatide and HISHS-2001 (Table 1, Fig. 5B), the biased activation of the GLP-1R results in increased amount of this receptor on the cell surfaces, while GIPR agonism result in proportionally higher GIPR internalization and thereby reduced receptor expression compared to the GLP-1R. In other terms, the receptor expression pattern and relationship could be an important factor for the success of GLP-1R-GIPR targeting type 2 diabetes treatment.
Another factor for addressing the studies of GLP-1R-GIPR co-activation and development of co-targeting drugs is the ratio of receptor activation e.g. the ratio of GLP-1 and GIP or the pharmacodynamics properties of a co-targeting compound towards each receptor. Taking pharmacodynamic properties of each receptor-ligand relationship, pharmacokinetic properties of the ligands as well as the factors of receptor expression highlighted above, and the presented shared signaling pathways into account, we leave this question unanswered. As described, numerous factors affect the implications of GLP-1R-GIPR coagonism and if the receptors are activated more than 50 % (above ligand EC 50 ), the ratio of receptor activation could likely be a minor factor.

Clinical GLP-1R-GIPR co-activation
As illustrated in Fig. 5A, the GLP-1R and GIPR share several intracellular signaling properties and in case of glucose-dependent insulin secretion in humans, the two hormones have been shown to act additively [2,3,21]. Presuming the mechanisms of action in the beta cells are common [119], the additive responses indicate a wide dynamic range of action in the beta cells of healthy individuals instead of parallel/distinct mechanisms. On top of that, the separate, unshared pathways (Figs. 3 and 4) may provide further basis for additive effects.
In a pathological state, such as type 2 diabetes, the beta cell responses to GLP-1 and GIP are however reduced, or in some cases, absent [120]. However, after a period of blood glucose normalization in patients with type 2 diabetes, the sensitivity and insulinotropic responses to GLP-1 and GIP are restored, indicating that the glucotoxicity affecting the beta cell function in general, are at least partly, reversible and that the intracellular machinery downstream of the GLP-1R and GIPR is intact [121,122]. Supporting the statement of intact GLP-1R, GIPR, and signaling properties in the beta cells in spite of the diabetic state, patients with the genetic diabetes variant (MODY3), where the insulin secretion is reduced, both GIP and GLP-1 administration results in increased insulin secretion [123].

GLP-1R-GIPR interactions
Within the evolving field of GLP-1R-GIPR co-targeting drugs, several levels of interaction can be considered. Both therapeutic compounds and endogenous ligands can interfere or interplay during ligand binding and receptor activation e.g. a compound can disrupt binding of endogenous ligands and act as an antagonist/partial agonist. As presented in this review, both G protein-and arrestin coupling can be affected e.g. by biased ligands and result in changed receptor expression and signaling ratios, pathways, and cellular responses. Co-targeting drugs are beneficial for administration, patient compliance, and drug approval processes, and overcome a possible risk of drug interactions, but we can, on the other hand, not adjust dosing separately and might reach thresholds for common pathways both for treatment and for adverse effects at a low dose than for each ligand separately. The discovery of the imbalanced profile of the GLP-1R-GIPR co-agonist tirzepatide, both on affinity towards the target receptors but most importantly also in the signaling profiles, illustrates the complexity of analysing co-targeting drugs [110,124,125].
Besides the receptor interactions presented above, GPCRs are able to interrelate and form either homodimers (constellation of the two identical GPCRs) or heterodimers (two different GPCRs) [126,127]. GLP-1Rs can dimerize [128][129][130], and GLP-1R and GIPR have been suggested to form heterodimers thereby affecting expression and signaling properties of each other [74]. In general, the consequences of the heterodimers are not known, but especially the functionality of the GLP-1R has been shown to be regulated by on cross-talk with the GIPR [130]. In light of the recent promising results on GLP-1-GIPR co-agonist, disruption of the dimerization could play a role for the clinical benefits of the GLP-1-GIPR co-agonist compounds.

GLP-1R agonism combined with GIPR agonism or antagonism
Presently, studies of both GIPR agonism and of GIPR antagonism combined with GLP-1R agonism have been shown to lead to the same end-point physiological effects such as weight loss and improved glucose tolerance [26,132,133]. Importantly, the compounds and methodologies used for preclinical studies are diverse and include e.g. the biased and imbalanced GLP-1-GIPR co-agonist tirzepatide, a GIPR antagonistic antibody, GIPR knockout mice, and GIP overexpression in mice [132,[134][135][136][137][138]. However, a comparison between a simple and non-selective GIPR agonist vs. a full GIPR antagonist in a clinical long-term setting has not been performed. Moreover, the signaling pathways leading to ligand-dependent biased signaling of both the GIPR and GLP-1R as well as the close association to GLP-1R activation, add complexity to the question of whether GIPR agonism or GIPR antagonism are suitable for treatment of metabolic diseases such as type 2 diabetes and obesity. Our present best answer is, that the ligand(s) that result in increased GLP-1R signaling relative to GIPR signaling, will be superior to GLP-1R mono-activation, however the molecular and cellular mechanisms remains to be fully described [130].
The concomitant GLP-1R activation is a common factor for the success of GIPR targeting treatments (agonist and antagonist) [107,[139][140][141] and the relative expression level/number of receptors present on the beta cell surface could be a key factor to understand the results of GLP-1-GIPR co-targeting treatments. However, the mechanism of the GLP-1R-GIPR interactions are, as presented in the review, not fully uncovered.

GLP-1R-GIPR co-agonism beyond the beta cells
Besides the insulinotropic actions, other main actions of GLP-1 administration are reduced appetite, delayed gastric emptying, and suppression of glucagon secretion from the pancreatic alpha cells [9,142]. Moreover, GIP administration result in reduced bone resorption, increased glucagon secretion (during hypo-and euglycemia), and increased lipid deposition in the adipose tissue [143,144] beyond its action on beta cells. Both GLP-1 and GIP administration also result in increased in heart rate [142,145], and could affect cardiac and vascular musculature [142,145]. GLP-1R-GIPR co-agonism will therefore not only affect the beta cells but several organs related to nutrient metabolism and glucose control. For most of the mentioned parameters, GLP-1R-GIPR co-agonism seem to have anti-diabetes and anti-obesity properties, however, especially glucagon secretion and appetite have, in acute studies, showed non-beneficial effects: In patients with type 2 diabetes, already treated with GLP-1R agonists, a GIP infusion resulted in higher glucagon secretion [145,146]. The presense of the GLP-1R on the human alpha cells has not been confirmed [147] but the GLP-1Rs are present on the delta cells and therefore glucagon-suppressing actions of GLP-1 can be via somatostatin [148]. The interplay between the GLP-1R and GIPR, which is present on the alpha cell [145], could therefore likely be affected be several extracellular factors as well as distinct receptor pharmacodynamics relations. For the other parameter 'appetite', persons with obesity had a higher food intake during infusion of the combination of GLP-1 and GIP than GLP-1 alone [150]. Fortunately, the clinical trials of the GLP-1R-GIPR co-agonist tirzepatide do not support the findings of higher glucagon levels and food intake [25] but whether this is due to the more complex islet interplay of GLP-1-induced glucagon suppression, the imbalanced and biased signaling profile of the specific compound or the physiological response to the long term exposure of GLP-1R-GIPR co-agonism remains to be clarified.

Conclusions
Via common signaling pathways in the beta cells, GLP-1R and GIPR activation result in insulin secretion and novel drug classes such as GLP-1R-GIPR co-agonists in drug development result in efficient antidiabetes treatment. The combination of GLP-1R agonism and GIPR antagonism does, however, also improve glucose tolerance and induce weight loss in preclinical and non-human primate studies. Based on in vitro pharmacological studies, the interplay of the two receptors and their intracellular signaling reveal that receptor internalization and selective signaling pathway activation form the basis of a complex but close relationship between the GLP-1R and GIPR. The synergistic actions seen using the biased and imbalanced GLP-1R and GIPR co-agonist tirzepatide, is an example of promising new drug classes that selectively modify the receptor expression and signaling pathways resulting in improved treatment.