PerspectiveType 2 Diabetes

GIP: No Longer the Neglected Incretin Twin?

See allHide authors and affiliations

Science Translational Medicine  15 Sep 2010:
Vol. 2, Issue 49, pp. 49ps47
DOI: 10.1126/scitranslmed.3001027


In the design of therapeutics to treat type 2 diabetes, researchers have exploited the observation that oral ingestion of nutrients leads to the secretion of glucose homeostasis–regulating incretin hormones (for example, glucagon-like-peptide–1) from the gut. Here, we discuss two recent papers that suggest that the “other” incretin hormone, gastric inhibitory polypeptide (GIP), also is important in the regulation of glucose homeostasis. These findings warrant further studies to unravel the mechanism of action of GIP in β-cells of the endocrine pancreas and to evaluate the possibility of designing novel therapeutics that target both incretin hormones.

Type 2 diabetes (T2D) has reached epidemic proportions and continues to impart a heavy economic burden on both developed and developing countries. T2D is characterized by an elevation in blood glucose—in the fasting state or 2 hours after an oral glucose load—that results from defects in insulin secretion on a background of insulin resistance. The 2-hour blood glucose concentration after an oral glucose challenge, which is a heritable quantitative trait, has been associated with T2D and also has been used to predict cardiovascular disease morbidity and mortality (1). Among the current therapies for the treatment of T2D, the use of the gut hormone glucagon-like-peptide–1 (GLP-1) (and its analogs, such as exendin-4) is based on a physiological increase in the gut peptide in response to ingested nutrients.

The concept that inter-organ communication regulates glucose homeostasis was suggested in the early 1900s, when it was recognized that certain factors secreted from the gut mucosa in response to nutrient intake are capable of stimulating the release of substances from the endocrine pancreas that reduce blood glucose concentrations (2). The term “incretin” was coined to refer to these glucose-lowering, intestinal-derived factors (3). It was later confirmed, using radioimmunoassay, that oral administration of glucose led to a greater increase in circulating insulin as compared with that of an equivalent dose of glucose given intravenously (4). This phenomenon was termed the “incretin effect” and is estimated to account for up to 70% of the total insulin secreted after oral glucose administration. Two recent reports (5, 6) have redirected the spotlight on incretin hormones and their receptors as potential candidate gene products involved in the development of T2D. Here, I discuss the implications of these findings with respect to this rampant metabolic disease and the intricate regulation of glucose homeostasis.


The incretin concept gained attention several decades ago with the discovery of the first such hormone, glucose-dependent insulinotropic polypeptide (GIP), which was originally named gastric inhibitory polypeptide on the basis of its ability to inhibit the secretion of gastric acid in dogs (7). However, studies using a purified preparation of GIP revealed that the polypeptide also stimulates insulin secretion in animals and humans. GIP was renamed after it was observed that its effects on gastric acid secretion occurred only at pharmacological doses as compared with GIP’s incretin action, which manifests at physiological concentrations.

It was later reported that GIP alone could not account for the full incretin effect. Thus, immunoneutralization of endogenous GIP activity attenuated but did not fully abolish the incretin effect in rodents, whereas in studies in humans, surgical resection of the ileum was associated with diminished incretin activity, despite the preservation of normal plasma GIP concentrations (8). The cloning and sequencing of the mammalian proglucagon gene led to the discovery of a second incretin hormone, GLP-1. In addition to glucagon, the proglucagon gene encodes the intestinal GLP-1 and GLP-2 hormones. Among the two glucagon-like peptides, only GLP-1 is able to stimulate insulin secretion from the pancreas.

The hormones GIP and GLP-1 are usually mentioned together when discussing incretin hormones and indeed share several common features. Both are secreted from the gut, with GIP being released from K cells in the small intestine primarily in response to fat or glucose ingestion and GLP-1 being secreted from intestinal L cells. Both are reported to potentiate glucose-stimulated insulin secretion in rodents and humans (Fig. 1) (810). Despite these similarities, considerably more focus in basic science laboratories and clinical arenas has centered on research into the action of GLP-1. The clinical focus was aided in part by the observation that T2D patients exhibit blunted incretin responses from the gut after nutrient ingestion, and the low amounts of GLP-1 in turn were unable to stimulate appropriate amounts of insulin release to maintain glucose within the physiological range (11). Consequently, GLP-1 has enjoyed prime status in terms of being harnessed for therapeutic applications for T2D (Fig. 1) (12).

Fig. 1. GIP versus GLP.

A comparison between GIP and GLP-1 in the regulation of β-cell function, β-cell mass, and peripheral effects in the overall maintenance of glucose homeostasis. GLP-1 affects glucose sensing by the pancreatic β-cell and suppresses glucagon secretion from pancreatic α-cells. GLP-1 also prolongs gastric emptying time and regulates appetite control in the brain. GIP enhances fatty acid synthesis by acting on white adipose tissue.



However, the recent observation by Saxena and colleagues (5) that variants at the GIP receptor (GIPR) gene locus are associated with elevated 2-hour glucose concentrations and decreased insulin secretion—and thus may contribute to T2D susceptibility—should generate enthusiasm for renewed efforts to decipher GIP’s effects on pancreatic β-cell biology and contribution to the maintenance of overall glucose homeostasis. The authors performed a meta-analysis combining 9 GWASs and replication stages for 29 independent genetic loci identified in 17 previous studies and identified 5 loci that were associated with elevated 2-hour glucose concentrations. Among the five loci [GIPR, VPS13C (vacuolar protein sorting homolog 13 C), ADCY5 (adenylate cyclase), GCKR (glucokinase regulator), and TCF7L2 (transcription factor 7-like 2)], the GIPR gene was newly associated with the 2-hour glucose measurement after adjusting for variations in age, sex, and body-mass index. Further, after adjusting for variations in fasting glucose levels, the authors observed an increased effect size in the association with 2-hour glucose concentrations that supported a specific role for the GIPR gene in the regulation of oral glucose disposal. Interestingly, the variants in the GIPR gene did not exhibit a genome-wide significant association with alterations in measurements of insulin resistance, which are also indicators of T2D. These findings reflect the difficulty and lack of power in such studies to uncover potential small physiological effects that can point to potential genome-wide associations between the variants and increased risk for developing T2D.

A defect in the release of incretin hormones in response to oral nutrients in T2D patients (11) and specifically, a defect in GIP action on insulin secretion were suggested by experimental results reported several years ago (13) that showed that GIPR is important for the promotion of insulin secretion after oral nutrient challenge. Indeed, in the current study (5) the authors assessed the association of GIPR gene variants with indices of insulin secretion after oral-glucose stimulation in humans from 13 studies and reported that the rs10423928 GIPR-A allele is associated with a reduction in early-phase insulin secretion (and thus, a lower-than-normal insulinogenic index). The lack of an association between the rs10423928 GIPR-A allele and the insulin response after an intravenous glucose tolerance test (IVGTT) underscores the importance of the gut and its incretin response in the observed association after the oral glucose tolerance test (GTT). These data are consistent with decade-old studies of GIPR gene knockout mice that correctly reported mild glucose intolerance and reduced insulin secretion after oral glucose challenge, but not in response to an intraperitoneal GTT (14), and underscore the importance of studies in genetically engineered mice.

The human GIPR gene variant that showed the most significant association with 2-hour oral GTT results (rs10423928) (5) contains an intronic single-nucleotide polymorphism (SNP) that exhibits strong nonrandom association (linkage disequilibrium) with another GIPR variant that has a missense mutation (rs1800437) that substitutes a glutamine residue for a glutamic acid (E354Q) at codon 354 (in exon 12). Previously, this latter mutation was considered to be a candidate for association with T2D, and in fact, one study reported that people homozygous for this Gln354-encoding allele had reduced fasting blood glucose levels and C-peptide (a product of proinsulin processing that is secreted along with insulin from the pancreatic β-cell) blood concentrations after oral glucose loading (15). These data suggest a role for GIPR in insulin secretion and are in agreement with the report by Saxena et al. (5). Nevertheless, a meta-analysis of 16 T2D genetic association studies showed that the rs10423928 GIPR-A allele was only one of those identified in (5) that was moderately associated with an increased risk for T2D. Together, these findings provide an example of how large GWASs might not be able to pinpont important candidate genes known to be involved in the pathophysiology of T2D because of small odds ratios or measure of effect size (16).


A direct role for the GIPR in modulating pancreatic β-cell secretion and mass is supported by the recent work of Renner and colleagues (6). To explore the role of GIPR in a large animal model, these authors used lentiviral transgenesis to create a transgenic pig that expresses a dominant-negative version of the GIPR (GIPRdn) in pancreatic islets. Similar to the observations in humans, the GIPRdn pigs exhibited poor tolerance to oral glucose challenge but not to intravenous glucose injection. This metabolic effect was associated with a striking decrease in β-cell proliferation in the pancreases of the GIPRdn pigs. Furthermore, both of these effects were exaggerated in older pigs (5-month-old piglets versus 1.25-year-old pigs). In contrast, the insulin secretion effects of exogenous exendin-4 (a GLP-1 analog) were significantly enhanced in the GIPRdn pigs, suggesting potential compensatory effects of the “other incretin,” GLP-1, when the GIPR is rendered inactive. Notably, no differences were detected in the levels of apoptosis of pancreatic β-cells between control and transgenic GIPRdn pigs. These apoptosis-related findings are in conflict with those reported in a recent paper by Widenmaier et al. (17), which showed that administration of a GIPR agonist to Vancouver diabetic fatty rats resulted in a suppression of apoptosis in pancreatic β-cells that in turn improved insulin responses and glycemic control. Whether this discrepancy in GIP action on β-cell apoptosis is species-dependent or is a feature of the truncated GIP analog (D-Ala2-GIP1-30) requires further investigation.


Although the two recent reports (5, 6) used two different approaches, together their findings (i) imply that humans are like pigs (at least with respect to glucose metabolism) and (ii) provide evidence for a role for GIP in the regulation of glucose homeostasis. However, several crucial questions remain.

Fig. 2. Double beats a single.

Shown here are the contrasting therapeutic potentials of single versus dual agonists that affect GIP and/or GLP-1 action in the regulation of glucose homeostasis and peripheral effects. (A) Altered GIP receptors affect glucose homeostasis by promoting defective insulin secretion and reduced pancreatic β-cell mass. A single GIP receptor agonist that can improve β-cell secretion and/or prevent β-cell death might lead to improved glycemic control. A potential deleterious effect of GIP agonists on white adipose cells can lead to increased adiposity. (B) A dual therapy approach perpetrated by a drug that affects both GIP and GLP-1 receptors enhances the activity of both hormones, particularly on β-cell mass and function. These changes may lead to improved glucose homeostasis. As part of a dual-agonist approach, the modulation of GLP-1 may have the effect of inhibiting glucagon secretion and reducing body weight. Adopting a GIP agonist that is selective for β-cells would circumvent effects on adipocytes.


First—and getting back to basics—unraveling of the mechanism (or mechanisms) of action of GIP in the regulation of both insulin secretion from pancreatic β-cells and β-cell mass will require further extensive investigation. Considering that both incretins have been reported to improve pancreatic β-cell function and mass, it will be important to carefully compare and contrast the functions of GIP and GLP-1 with a view to identifying common and unique proteins activated by their respective signaling pathways. Further, although GLP-1 has been reported to act favorably on other organs (for example, gut and brain) to maintain glucose homeostasis in humans (Fig. 1) (12), the effects of GIP on adipose tissue have been considered to promote obesity.

To this end, it is worth considering the lessons learned from rodent models. In contrast to the β-cell defects observed in both people and pigs that carry mutant versions of the GIPR gene, mice in which the GIPR gene has been knocked out exhibit only mild glucose intolerance (18) that probably results from some compensatory protective mechanisms by GLP-1 signaling. This hypothesis stems from the observation that mice with a double knockout of the genes that encode the GLP-1 and GIP receptors [double incretin receptor knockout (DIRKO) mice], (19) show glucose intolerance that is slightly more severe compared with that displayed by the individual knockout mice (20, 21); DIRKO mice have normal body weight and normal glucagon secretion and display defects in glucose-stimulated insulin secretion in response to glucose received orally but not by intraperitoneal injection. Also, DIRKO mice do not develop overt diabetes.

It may be that, in rodents, the incretin pathway is not as critical to the regulation of glucose homeostasis as it is in pigs or people and that other as-yet unexplored pathways compensate in the absence of the two incretin receptors (22). It would be important to understand whether growth factor pathways that are known to mediate β-cell proliferation and survival (23) by interacting functionally with GLP-1R (2426) exhibit similar interactions with the GIPR (Fig. 1) (27, 28). Another approach would be to express the GIPR variant in human islet/β-cells and/or create a mouse model that expresses the variant in a β-cell–specific manner so as to explore the consequences for glucose homeostasis—a translational strategy that has been successfully adopted for investigating polymorphisms in the tribbles gene, which encodes a protein that affects β-cell function in humans (29). The actions of GLP-1R and GIPR in human islet cells have not been fully explored and thus are not well documented; this is because of the absence of suitable methods to directly assess β-cell growth in in vivo studies in humans and because of the limited availability of good-quality, freshly isolated human islets for in vitro experiments. These technical problems notwithstanding, efforts to understand the actions of GIP in human islets/β-cells are crucial if we are to improve the targeting strategy aimed at enhancing the function of the two incretins in humans with T2D.

Second, considering that both the GIP and GLP-1 hormones are released in a physiological context, the possibility of developing dual agonists that can affect both GIP and GLP-1 receptors is worth exploring, as has been done for other targets (Fig. 2) (30, 31). Several GLP-1 agonists, GLP-1 receptor agonists, and molecules that enhance the endogenous amounts of GLP-1 are currently either in clinical use or in various phases of clinical development for the treatment of T2D and provide a good starting point for the discovery or fashioning of dual agonists (32). Among peptides known to enhance incretin effects, it may be useful to examine Xenin-25, which was isolated from a subset of K cells from the gut and has been reported to potentiate GIP-mediated insulin release by activating nonganglionic cholinergic neurons as part of an enteric-neuronal-pancreatic pathway (33).

A dual agonist that has incretin-like properties would be desirable at the level of regulating β-cell function and mass. In addition to the influence of GLP-1 agonists on the endocrine pancreas, effects on gastric emptying, appetite control, and body weight maintenance would be an important advantage in the dual-agonist approach (Fig. 2). However, the potential deleterious effects of GIP agonists on fat metabolism should be borne in mind. For example, GIP has been suggested to act as a pro-obesity hormone (34, 35) by promoting lipogenesis in the presence of insulin; GIP achieves this effect by up-regulating transcription of the human adipocyte lipoprotein lipase (LPL) gene via a pathway involving the cyclic adenosine monophosphate (cAMP) response element–binding protein (CREB) and the cAMP-responsive CREB co-activator 2–mediated pathway (36). Several studies indicate that reduced insulin responses to GIP and elimination of GIP signaling enhances resistance to obesity in rodents (34, 3739). These reports argue that GIPR agonists would be poor drugs because of their potential pro-obesity effects in T2D patients. Although these studies urge caution in choosing GIPR agonists for the treatment of T2D, McIntosh and colleagues have recently reported on a GIPR agonist that displays selective physiological effects on β-cell function without enhancing obesity in rodents (17). It is worth following up and building on these interesting observations so as to explore whether these agonists have similar properties in human tissue.


Lastly, one aspect that has not been directly tested in human studies and is important from an evolutionary perspective is the importance of transgenerational effects in which DNA sequence variations (for example, in the GIPR gene) in one generation affect phenotypes in subsequent generations without inheritance of the original genetic variant (Fig. 3) (40, 41). These transgenerational phenotypic effects, which perhaps contribute to “missing heritability,” are thought to be caused by epigenetic changes in the parent that persist in offspring. Examples of this phenomenon in mammals that have overt phenotypic consequences include (i) the cardiac hypertrophy that occurs in the offspring of mice that were injected with microRNAs that block expression of the cdk9 gene, a key regulator of cardiac growth (42), and (ii) mutations in the Dnd1 gene that promote testicular cancer in mice and their offspring as a consequence of transgenerational epistasis, in which interactions occur between an inherited epigenetic state in the parent and conventional genetic variants in the offspring (43).

Fig. 3. Possible persistence.

Depicted here is the hypothesis that DNA sequence variations in the gene that encodes the GIPR give rise to transgenerational phenotypic effects that result from epigenetic changes (for example, those involving microRNAs/RNA silencing, DNA methylation, and histone modifications) in the parent that persist in children and grandchildren, even if the offspring do not carry the original variation in the GIPR gene sequence. Also possible is that the phenotypes can become weaker, drift, or exhibit reversion in offspring. The precise mechanisms that underlie the evolution of the phenotypic effects in human subjects warrant further investigation.


A major challenge to the discovery of potential transgenerational effects in humans is controlling the influence of background genetic variation, environmental triggers, and social cues. It remains to be seen whether the frequency of the affected traits and strength of the phenotypic effects of individuals that express the GIPR gene variants is altered in the next generation. This speculation, however remote, can be confounding and reveals the limitations of current GWAS studies, which are not designed to directly consider epigenetic individual- or population-specific variations. Thus it will be important for genetic epidemiologists to develop studies to monitor and model these changes in humans with a view to maximizing effective therapy across generations.


  • Citation: R. N. Kulkarni, GIP: No longer the neglected incretin twin? Sci. Transl. Med. 2, 49ps47 (2010).

References and Notes

  1. Acknowledgments: I thank K. Parlee for excellent assistance with manuscript preparation and gratefully acknowledge support from the U.S. National Institutes of Health (grant DK RO1 67536) and the American Diabetes Association. Competing interests: The author declares no competing interests.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article