The incretin co-agonist tirzepatide requires GIPR for hormone secretion from human islets.

The incretins glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide 1 (GLP-1) mediate insulin responses that are proportionate to nutrient intake to facilitate glucose tolerance1. The GLP-1 receptor (GLP-1R) is an established drug target for the treatment of diabetes and obesity2, whereas the therapeutic potential of the GIP receptor (GIPR) is a subject of debate. Tirzepatide is an agonist at both the GIPR and GLP-1R and is a highly effective treatment for type 2 diabetes and obesity3,4. However, although tirzepatide activates GIPR in cell lines and mouse models, it is not clear whether or how dual agonism contributes to its therapeutic benefit. Islet beta cells express both the GLP-1R and the GIPR, and insulin secretion is an established mechanism by which incretin agonists improve glycemic control5. Here, we show that in mouse islets, tirzepatide stimulates insulin secretion predominantly through the GLP-1R, owing to reduced potency at the mouse GIPR. However, in human islets, antagonizing GIPR activity consistently decreases the insulin response to tirzepatide. Moreover, tirzepatide enhances glucagon secretion and somatostatin secretion in human islets. These data demonstrate that tirzepatide stimulates islet hormone secretion from human islets through both incretin receptors.

Discovery of a potent GIPR peptide antagonist that is effective in rodent and human systems.

OBJECTIVE: Glucose-dependent insulinotropic polypeptide (GIP) is one of the two major incretin factors that regulate metabolic homeostasis. Genetic ablation of its receptor (GIPR) in mice confers protection against diet-induced obesity (DIO), while GIPR neutralizing antibodies produce additive weight reduction when combined with GLP-1R agonists in preclinical models and clinical trials. Conversely, GIPR agonists have been shown to promote weight loss in rodents, while dual GLP-1R/GIPR agonists have proven superior to GLP-1R monoagonists for weight reduction in clinical trials. We sought to develop a long-acting, specific GIPR peptide antagonist as a tool compound suitable for investigating GIPR pharmacology in both rodent and human systems. METHODS: We report a structure-activity relationship of GIPR peptide antagonists based on the human and mouse GIP sequences with fatty acid-based protraction. We assessed these compounds in vitro, in vivo in DIO mice, and ex vivo in islets from human donors. RESULTS: We report the discovery of a GIP(5-31) palmitoylated analogue, [Nα-Ac, L14, R18, E21] hGIP(5-31)-K11 (γE-C16), which potently inhibits in vitro GIP-mediated cAMP generation at both the hGIPR and mGIPR. In vivo, this peptide effectively blocks GIP-mediated reductions in glycemia in response to exogenous and endogenous GIP and displays a circulating pharmacokinetic profile amenable for once-daily dosing in rodents. Co-administration with the GLP-1R agonist semaglutide and this GIPR peptide antagonist potentiates weight loss compared to semaglutide alone. Finally, this antagonist inhibits GIP- but not GLP-1-stimulated insulin secretion in intact human islets. CONCLUSIONS: Our work demonstrates the discovery of a potent, specific, and long-acting GIPR peptide antagonist that effectively blocks GIP action in vitro, ex vivo in human islets, and in vivo in mice while producing additive weight-loss when combined with a GLP-1R agonist in DIO mice.

Next generation GLP-1/GIP/glucagon triple agonists normalize body weight in obese mice.

OBJECTIVE: Pharmacological strategies that engage multiple mechanisms-of-action have demonstrated synergistic benefits for metabolic disease in preclinical models. One approach, concurrent activation of the glucagon-like peptide-1 (GLP-1), glucose-dependent insulinotropic peptide (GIP), and glucagon (Gcg) receptors (i.e. triagonism), combines the anorectic and insulinotropic activities of GLP-1 and GIP with the energy expenditure effect of glucagon. While the efficacy of triagonism in preclinical models is known, the relative contribution of GcgR activation remains unassessed. This work aims to addresses that central question. METHODS: Herein, we detail the design of unimolecular peptide triagonists with an empirically optimized receptor potency ratio. These optimized peptide triagonists employ a protraction strategy permitting once-weekly human dosing. Additionally, we assess the effects of these peptides on weight-reduction, food intake, glucose control, and energy expenditure in an established DIO mouse model compared to clinically relevant GLP-1R agonists (e.g. semaglutide) and dual GLP-1R/GIPR agonists (e.g. tirzepatide). RESULTS: Optimized triagonists normalize body weight in DIO mice and enhance energy expenditure in a manner superior to that of GLP-1R mono-agonists and GLP-1R/GIPR co-agonists. CONCLUSIONS: These pre-clinical data suggest unimolecular poly-pharmacology as an effective means to target multiple mechanisms contributing to obesity and further implicate GcgR activation as the differentiating factor between incretin receptor mono- or dual-agonists and triagonists.

Optimization of Truncated Glucagon Peptides to Achieve Selective, High Potency, Full Antagonists.

Antagonism of glucagon's biological action is a proven strategy for decreasing glucose in diabetic animals and patients. To achieve full, potent, and selective suppression, we chemically optimized N-terminally truncated glucagon fragments for the identification and establishment of the minimum sequence peptide, [Glu9]glucagon(6-29) amide (11) as a full antagonist in cellular signaling and receptor binding (IC50 = 36 nM). Substitution of Phe6 with l-3-phenyllactic acid (Pla) produced [Pla6, Glu9]glucagon(6-29) amide (21), resulting in a 3-fold improvement in receptor binding (IC50 = 12 nM) and enhanced antagonist potency. Further substitution of Glu9 and Asn28 with aspartic acid yielded [Pla6, Asp28]glucagon amide (26), which demonstrated a further increase in inhibitory potency (IC50 = 9 nM), and improved aqueous solubility. Peptide 26 and a palmitoylated analogue, [Pla6, Lys10(γGluγGlu-C16), Asp28]glucagon(6-29) amide (31), displayed sustained duration in vivo action that successfully reversed glucagon-induced glucose elevation in mice.

The glucose-dependent insulinotropic polypeptide (GIP) regulates body weight and food intake via CNS-GIPR signaling.

Uncertainty exists as to whether the glucose-dependent insulinotropic polypeptide receptor (GIPR) should be activated or inhibited for the treatment of obesity. Gipr was recently demonstrated in hypothalamic feeding centers, but the physiological relevance of CNS Gipr remains unknown. Here we show that HFD-fed CNS-Gipr KO mice and humanized (h)GIPR knockin mice with CNS-hGIPR deletion show decreased body weight and improved glucose metabolism. In DIO mice, acute central and peripheral administration of acyl-GIP increases cFos neuronal activity in hypothalamic feeding centers, and this coincides with decreased body weight and food intake and improved glucose handling. Chronic central and peripheral administration of acyl-GIP lowers body weight and food intake in wild-type mice, but shows blunted/absent efficacy in CNS-Gipr KO mice. Also, the superior metabolic effect of GLP-1/GIP co-agonism relative to GLP-1 is extinguished in CNS-Gipr KO mice. Our data hence establish a key role of CNS Gipr for control of energy metabolism.

Selection and progression of unimolecular agonists at the GIP, GLP-1, and glucagon receptors as drug candidates.

The continued global growth in the prevalence of obesity coupled with the limited number of efficacious and safe treatment options elevates the importance of innovative pharmaceutical approaches. Combinatorial strategies that harness the metabolic benefits of multiple hormonal mechanisms have emerged at the preclinical and more recently clinical stages of drug development. A priority has been anti-obesity unimolecular peptides that function as balanced, high potency poly-agonists at two or all the cellular receptors for the endocrine hormones glucagon-like peptide-1 (GLP-1), glucose-dependent insulinotropic polypeptide (GIP), and glucagon. This report reviews recent progress in this area, with emphasis on what the initial clinical results demonstrate and what remains to be addressed.

Insulin-like peptide 5 fails to improve metabolism or body weight in obese mice.

Insulin-like peptide 5 (INSL5) is a member of the insulin-like family of peptides. It has been reported to be orexigenic in rodent models of obesity with impaired glucose metabolism. We attempted to confirm this property as a first step in establishing the ability of INSL5 to successfully integrate with other agents more proven in their ability to reverse obesity and improve metabolism. INSL5 was chemically synthesized by two alternative methods to a native form and one that was site-specifically conjugated to a 20 KDa polyethylene glycol (PEG) polymer. The pharmacology of each peptide was assessed by high-dose chronic administration in normal and obese mice. INSL5 failed to produce pharmacologically relevant effects on food intake, body weight or glucose control indicative of a negligible role of the peptide in the control of feeding and glucose metabolism.

Optimized GIP analogs promote body weight lowering in mice through GIPR agonism not antagonism.

OBJECTIVE: Structurally-improved GIP analogs were developed to determine precisely whether GIP receptor (GIPR) agonism or antagonism lowers body weight in obese mice. METHODS: A series of peptide-based GIP analogs, including structurally diverse agonists and a long-acting antagonist, were generated and characterized in vitro using functional assays in cell systems overexpressing human and mouse derived receptors. These analogs were characterized in vivo in DIO mice following acute dosing for effects on glycemic control, and following chronic dosing for effects on body weight and food intake. Pair-feeding studies and indirect calorimetry were used to survey the mechanism for body weight lowering. Congenital Gipr-/- and Glp1r-/- DIO mice were used to investigate the selectivity of the agonists and to ascribe the pharmacology to effects mediated by the GIPR. RESULTS: Non-acylated, Aib2 substituted analogs derived from human GIP sequence showed full in vitro potency at human GIPR and subtly reduced in vitro potency at mouse GIPR without cross-reactivity at GLP-1R. These GIPR agonists lowered acute blood glucose in wild-type and Glp1r-/- mice, and this effect was absent in Gipr-/- mice, which confirmed selectivity towards GIPR. Chronic treatment of DIO mice resulted in modest yet consistent, dose-dependent decreased body weight across many studies with diverse analogs. The mechanism for body weight lowering is due to reductions in food intake, not energy expenditure, as suggested by pair-feeding studies and indirect calorimetry assessment. The weight lowering effect was preserved in DIO Glp-1r-/- mice and absent in DIO Gipr-/- mice. The body weight lowering efficacy of GIPR agonists was enhanced with analogs that exhibit higher mouse GIPR potency, with increased frequency of administration, and with fatty-acylated peptides of extended duration of action. Additionally, a fatty-acylated, N-terminally truncated GIP analog was shown to have high in vitro antagonism potency for human and mouse GIPR without cross-reactive activity at mouse GLP-1R or mouse glucagon receptor (GcgR). This acylated antagonist sufficiently inhibited the acute effects of GIP to improve glucose tolerance in DIO mice. Chronic treatment of DIO mice with high doses of this acylated GIPR antagonist did not result in body weight change. Further, co-treatment of this acylated GIPR antagonist with liraglutide, an acylated GLP-1R agonist, to DIO mice did not result in increased body weight lowering relative to liraglutide-treated mice. Enhanced body weight lowering in DIO mice was evident however following co-treatment of long-acting selective individual agonists for GLP-1R and GIPR, consistent with previous data. CONCLUSIONS: We conclude that peptide-based GIPR agonists, not peptide-based GIPR antagonists, that are suitably optimized for receptor selectivity, cross-species activity, and duration of action consistently lower body weight in DIO mice, although with moderate efficacy relative to GLP-1R agonists. These preclinical rodent pharmacology results, in accordance with recent clinical results, provide definitive proof that systemic GIPR agonism, not antagonism, is beneficial for body weight loss.

Hepatic Glucagon Receptor Signaling Enhances Insulin-Stimulated Glucose Disposal in Rodents.

Glucagon receptor (GCGR) agonists cause hyperglycemia but also weight loss. However, GCG-like peptide 1 receptor (GLP1R)/GCGR mixed agonists do not exhibit the diabetogenic effects often attributed to GCGR activity. Thus, we sought to investigate the effect of glucagon agonism on insulin action and glucose homeostasis. Acute GCGR agonism induced immediate hyperglycemia, followed by improved glucose tolerance and enhanced glucose-stimulated insulin secretion. Moreover, acute GCGR agonism improved insulin tolerance in a dose-dependent manner in both lean and obese mice. Improved insulin tolerance was independent of GLP1R, FGF21, and hepatic glycogenolysis. Moreover, we observed increased glucose infusion rate, disposal, uptake, and suppressed endogenous glucose production during euglycemic clamps. Mice treated with insulin and GCGR agonist had enhanced phosphorylation of hepatic AKT at Ser473; this effect was reproduced in isolated mouse primary hepatocytes and resulted in increased AKT kinase activity. These data reveal that GCGR agonism enhances glucose tolerance, in part, by augmenting insulin action, with implications for the use of GCGR agonism in therapeutic strategies for diabetes.

Glucagon Receptor Signaling Regulates Energy Metabolism via Hepatic Farnesoid X Receptor and Fibroblast Growth Factor 21.

Glucagon, an essential regulator of glucose and lipid metabolism, also promotes weight loss, in part through potentiation of fibroblast growth factor 21 (FGF21) secretion. However, FGF21 is only a partial mediator of metabolic actions ensuing from glucagon receptor (GCGR) activation, prompting us to search for additional pathways. Intriguingly, chronic GCGR agonism increases plasma bile acid levels. We hypothesized that GCGR agonism regulates energy metabolism, at least in part, through farnesoid X receptor (FXR). To test this hypothesis, we studied whole-body and liver-specific FXR-knockout (Fxr∆liver) mice. Chronic GCGR agonist (IUB288) administration in diet-induced obese (DIO) Gcgr, Fgf21, and Fxr whole-body or liver-specific knockout (∆liver) mice failed to reduce body weight when compared with wild-type (WT) mice. IUB288 increased energy expenditure and respiration in DIO WT mice, but not Fxr∆liver mice. GCGR agonism increased [14C]palmitate oxidation in hepatocytes isolated from WT mice in a dose-dependent manner, an effect blunted in hepatocytes from Fxr∆liver mice. Our data clearly demonstrate that control of whole-body energy expenditure by GCGR agonism requires intact FXR signaling in the liver. This heretofore-unappreciated aspect of glucagon biology has implications for the use of GCGR agonism in the therapy of metabolic disorders.

Molecular elements in FGF19 and FGF21 defining KLB/FGFR activity and specificity.

OBJECTIVE: To signal, FGF19 and FGF21 require co-receptor ßKlotho (KLB) to act in concert with FGF receptors, and yet there is appreciable variance in the C-terminal sequences of these two novel metabolic hormones where binding is believed to be primary. We seek to determine the functional consequences for these amino acid differences and determine whether such information can be used to design high potency antagonists and agonists. METHODS: We employed a functional in vitro assay to identify C-terminal protein fragments capable of fully blocking KLB-mediated FGF19 and 21 receptor signaling. The key residues in each hormone responsible for support full bioactivity were identified through peptide-based Ala-scanning. Chemical optimization of the peptides was employed to increase their antagonistic potency. An optimized sequence as a substituted part of a full length FGF21 was assessed for enhanced FGFR/KLB-mediated agonism using tissue culture and obese mice. RESULTS: C-terminal FGF19 and FGF21 peptides of relatively short length were observed to potently inhibit the activity of these two hormones, in vitro and in vivo. These FGFs of different sequence also demonstrated a striking conservation of structural determinants to maintain KLB binding. A single C-terminal amino acid in FGF19 was observed to modulate relative activity through FGFR1 and FGFR4. The substitution of native FGF21 C-terminal sequence with a peptide optimized for the highest antagonistic activity resulted in significantly enhanced FGF potency, as measured by in vitro signaling and improvements in metabolic outcomes in diet-induced obese mice. CONCLUSIONS: We report here the ability of short C-terminal peptides to bind KLB and function as antagonists of FGF19 and 21 actions. These proteins maintain high conservation of sequence in those residues central to KLB binding. An FGF21 chimeric protein possessing an optimized C-terminal sequence proved to be a super-agonist in delivery of beneficial metabolic effects in obese mice.

Desacyl Ghrelin Decreases Anxiety-like Behavior in Male Mice.

Ghrelin is a 28-amino acid polypeptide that regulates feeding, glucose metabolism, and emotionality (stress, anxiety, and depression). Plasma ghrelin circulates as desacyl ghrelin (DAG) or, in an acylated form, acyl ghrelin (AG), through the actions of ghrelin O-acyltransferase (GOAT), exhibiting low or high affinity, respectively, for the growth hormone secretagogue receptor (GHSR) 1a. We investigated the role of endogenous AG, DAG, and GHSR1a signaling on anxiety and stress responses using ghrelin knockout (Ghr KO), GOAT KO, and Ghsr stop-floxed (Ghsr null) mice. Behavioral and hormonal responses were tested in the elevated plus maze and light/dark (LD) box. Mice lacking both AG and DAG (Ghr KO) increased anxiety-like behaviors across tests, whereas anxiety reactions were attenuated in DAG-treated Ghr KO mice and in mice lacking AG (GOAT KO). Notably, loss of GHSR1a (Ghsr null) did not affect anxiety-like behavior in any test. Administration of AG and DAG to Ghr KO mice with lifelong ghrelin deficiency reduced anxiety-like behavior and decreased phospho-extracellular signal-regulated kinase phosphorylation in the Edinger-Westphal nucleus in wild-type mice, a site normally expressing GHSR1a and involved in stress- and anxiety-related behavior. Collectively, our data demonstrate distinct roles for endogenous AG and DAG in regulation of anxiety responses and suggest that the behavioral impact of ghrelin may be context dependent.

Fibroblast activation protein (FAP) as a novel metabolic target.

OBJECTIVE: Fibroblast activation protein (FAP) is a serine protease belonging to a S9B prolyl oligopeptidase subfamily. This enzyme has been implicated in cancer development and recently reported to regulate degradation of FGF21, a potent metabolic hormone. Using a known FAP inhibitor, talabostat (TB), we explored the impact of FAP inhibition on metabolic regulation in mice. METHODS: To address this question we evaluated the pharmacology of TB in various mouse models including those deficient in FGF21, GLP1 and GIP signaling. We also studied the ability of FAP to process FGF21 in vitro and TB to block FAP enzymatic activity. RESULTS: TB administration to diet-induced obese (DIO) animals led to profound decreases in body weight, reduced food consumption and adiposity, increased energy expenditure, improved glucose tolerance and insulin sensitivity, and lowered cholesterol levels. Total and intact plasma FGF21 were observed to be elevated in TB-treated DIO mice but not lean animals where the metabolic impact of TB was significantly attenuated. Furthermore, and in stark contrast to naïve DIO mice, the administration of TB to obese FGF21 knockout animals demonstrated no appreciable effect on body weight or any other measures of metabolism. In support of these results we observed no enzymatic degradation of human FGF21 at either end of the protein when FAP was inhibited in vitro by TB. CONCLUSIONS: We conclude that pharmacological inhibition of FAP enhances levels of FGF21 in obese mice to provide robust metabolic benefits not observed in lean animals, thus validating this enzyme as a novel drug target for the treatment of obesity and diabetes.

Chemical Hybridization of Glucagon and Thyroid Hormone Optimizes Therapeutic Impact for Metabolic Disease.

Glucagon and thyroid hormone (T3) exhibit therapeutic potential for metabolic disease but also exhibit undesired effects. We achieved synergistic effects of these two hormones and mitigation of their adverse effects by engineering chemical conjugates enabling delivery of both activities within one precisely targeted molecule. Coordinated glucagon and T3 actions synergize to correct hyperlipidemia, steatohepatitis, atherosclerosis, glucose intolerance, and obesity in metabolically compromised mice. We demonstrate that each hormonal constituent mutually enriches cellular processes in hepatocytes and adipocytes via enhanced hepatic cholesterol metabolism and white fat browning. Synchronized signaling driven by glucagon and T3 reciprocally minimizes the inherent harmful effects of each hormone. Liver-directed T3 action offsets the diabetogenic liability of glucagon, and glucagon-mediated delivery spares the cardiovascular system from adverse T3 action. Our findings support the therapeutic utility of integrating these hormones into a single molecular entity that offers unique potential for treatment of obesity, type 2 diabetes, and cardiovascular disease.

Pyridyl-alanine as a Hydrophilic, Aromatic Element in Peptide Structural Optimization.

Glucagon (Gcg) 1 serves a seminal physiological role in buffering against hypoglycemia, but its poor biophysical properties severely complicate its medicinal use. We report a series of novel glucagon analogues of enhanced aqueous solubility and stability at neutral pH, anchored by Gcg[Aib16]. Incorporation of 3- and 4-pyridyl-alanine (3-Pal and 4-Pal) enhanced aqueous solubility of glucagon while maintaining biological properties. Relative to native hormone, analogue 9 (Gcg[3-Pal6,10,13, Aib16]) demonstrated superior biophysical character, better suitability for medicinal purposes, and comparable pharmacology against insulin-induced hypoglycemia in rats and pigs. Our data indicate that Pal is a versatile surrogate to natural aromatic amino acids and can be employed as an alternative or supplement with isoelectric adjustment to refine the biophysical character of peptide drug candidates.

Unimolecular Polypharmacy for Treatment of Diabetes and Obesity.

Many complex diseases have historically proven to be defiant to the best mono-therapeutic approaches. Several examples of combination therapies have largely overcome such challenges, notably for the treatment of severe hypertension and tuberculosis. Obesity and its consequences, such as type 2 diabetes, have proven to be equally resistant to therapeutic approaches based on single medicines. Proper management of type 2 diabetes often requires adjunctive medications, and the recent registration of a few compound mixtures has set the precedent for combinatorial treatment of obesity. On the other hand, double or triple therapeutic combinations are more difficult to advance to regulatory approval than single molecules. More recently, several classes of novel unimolecular combination therapeutics have emerged with superior efficacy than currently prescribed options and pose the potential to reverse obesity and type 2 diabetes. Here, we summarize the discovery, pre-clinical validation, and first clinical test of such peptide hormone poly-agonist drug candidates.

Reappraisal of GIP Pharmacology for Metabolic Diseases.

Glucagon-like peptide-1 (GLP-1) analogs are considered the best current medicines for type 2 diabetes (T2D) and obesity due to their actions in lowering blood glucose and body weight. Despite similarities to GLP-1, glucose-dependent insulinotropic polypeptide (GIP) has not been extensively pursued as a medical treatment for T2D. This is largely based on observations of diminished responses of GIP to lower blood glucose in select patients, as well as evidence from rodent knockout models implying that GIP promotes obesity. These findings have prompted the belief in some, that inhibiting GIP action might be beneficial for metabolic diseases. However, a growing body of new evidence - including data based on refined genetically modified models and improved pharmacological agents - suggests a paradigm shift on how the GIP system should be manipulated for metabolic benefits.

A rationally designed monomeric peptide triagonist corrects obesity and diabetes in rodents.

We report the discovery of a new monomeric peptide that reduces body weight and diabetic complications in rodent models of obesity by acting as an agonist at three key metabolically-related peptide hormone receptors: glucagon-like peptide-1 (GLP-1), glucose-dependent insulinotropic polypeptide (GIP) and glucagon receptors. This triple agonist demonstrates supraphysiological potency and equally aligned constituent activities at each receptor, all without cross-reactivity at other related receptors. Such balanced unimolecular triple agonism proved superior to any existing dual coagonists and best-in-class monoagonists to reduce body weight, enhance glycemic control and reverse hepatic steatosis in relevant rodent models. Various loss-of-function models, including genetic knockout, pharmacological blockade and selective chemical knockout, confirmed contributions of each constituent activity in vivo. We demonstrate that these individual constituent activities harmonize to govern the overall metabolic efficacy, which predominantly results from synergistic glucagon action to increase energy expenditure, GLP-1 action to reduce caloric intake and improve glucose control, and GIP action to potentiate the incretin effect and buffer against the diabetogenic effect of inherent glucagon activity. These preclinical studies suggest that, so far, this unimolecular, polypharmaceutical strategy has potential to be the most effective pharmacological approach to reversing obesity and related metabolic disorders.

The role of ß cell glucagon-like peptide-1 signaling in glucose regulation and response to diabetes drugs.

Glucagon-like peptide-1 (GLP-1), an insulinotropic gut peptide released after eating, is essential for normal glucose tolerance (GT). To determine whether this effect is mediated directly by GLP-1 receptors (GLP1R) on islet ß cells, we developed mice with ß cell-specific knockdown of Glp1r. ß cell Glp1r knockdown mice had impaired GT after intraperitoneal (i.p.) glucose and did not secrete insulin in response to i.p. or intravenous GLP-1. However, they had normal GT after oral glucose, a response that was impaired by a GLP1R antagonist. ß cell Glp1r knockdown mice had blunted responses to a GLP1R agonist but intact glucose lowering with a dipeptidylpeptidase 4 (DPP-4) inhibitor. Thus, in mice, ß cell Glp1rs are required to respond to hyperglycemia and exogenous GLP-1, but other factors compensate for reduced GLP-1 action during meals. These results support a role for extraislet GLP1R in oral glucose tolerance and paracrine regulation of ß cells by islet GLP-1.