| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
FRONTIERS
Departments of 1Physiology and 2Medicine, University of Toronto, Toronto, Ontario, Canada
Submitted 7 March 2007 ; accepted in final form 13 April 2007
ABSTRACT
Glucagon-like peptide-2 (GLP-2) is a pleiotropic hormone that affects multiple facets of intestinal physiology, including growth, barrier function, digestion, absorption, motility, and blood flow. The mechanisms through which GLP-2 produces these actions are complex, involving unique signaling mechanisms and multiple indirect mediators. As clinical trials have begun for the use of GLP-2 in a variety of intestinal disorders, the elucidation of such mechanisms is vital. The GLP-2 receptor (GLP-2R) is a G protein-coupled receptor, signaling through multiple G proteins to affect the cAMP and mitogen-activated protein kinase pathways, leading to both proliferative and antiapoptotic cellular responses. The GLP-2R also demonstrates unique mechanisms for receptor trafficking. Expression of the GLP-2R in discrete sets of intestinal cells, including endocrine cells, subepithelial myofibroblasts, and enteric neurons, has led to the hypothesis that GLP-2 acts indirectly through multiple mediators to produce its biological effects. Indeed, several studies have now provided important mechanistic data illustrating several of the indirect pathways of GLP-2 action. Thus, insulin-like growth factor I has been demonstrated to be required for GLP-2-induced crypt cell proliferation, likely involving activation of 
-catenin signaling. Furthermore, vasoactive intestinal polypeptide modulates the actions of GLP-2 in models of intestinal inflammation, while keratinocyte growth factor is required for GLP-2-induced colonic mucosal growth and mucin expression. Finally, enteric neural GLP-2R signaling affects intestinal blood flow through a nitric oxide-dependent mechanism. Determining how GLP-2 produces its full range of biological effects, which mediators are involved, and how these mediators interact is a continuing area of active research.
growth; intestine; receptor; signaling
GLP-2 Is a Multifaceted Intestinal Growth Factor
The observations that GLP-2 is a potent intestinal mitogen in rodents and possesses therapeutic potential in humans, have led to considerable interest in the mechanisms through which this hormone regulates intestinal epithelial growth (13). GLP-2 is a 33-amino acid peptide and, along with its cognate hormone GLP-1, is liberated by prohormone convertase-1/3-mediated cleavage of proglucagon in the intestinal endocrine L cell. GLP-2 is secreted in response to nutrient intake and is subsequently inactivated by dipeptidyl peptidase IV cleavage and cleared by the kidney, conferring a relatively short biological half-life of 
7 min. Consequently, the use of native GLP-2 has been superceded by a degradation-resistant analog, (Gly2)GLP-2 (Teduglutide), in animal research models and human clinical trials. Such studies have revealed that GLP-2 is a pleiotropic hormone, affecting multiple facets of intestinal physiology. Foremost among these is the ability of GLP-2 to increase small and large intestinal weight through stimulation of epithelial cell proliferation and inhibition of apoptosis, leading to enlarged crypts and villi and, hence, an enhanced absorptive surface area. In fact, a physiological role for GLP-2 appears to be the restoration of epithelial growth following periods of fasting (38). GLP-2 also increases the capacity for carbohydrate, amino acid, and lipid absorption and increases the activity of epithelial brush-border digestive enzymes and nutrient transporters. Epithelial barrier capacity is enhanced by GLP-2 through decreases in transcellular and paracellular permeability, as well as accelerated wound closure following injury. In addition to its effects on the epithelium, GLP-2 also stimulates intestinal blood flow and inhibits gastrointestinal motility. Excitingly, GLP-2 has recently been shown to produce anti-inflammatory effects, independent of its proliferative actions (39). However, the growth-promoting actions of GLP-2 appear to be context -specific and are dependent on the developmental state and health of the intestine. For instance, in contrast to the mainly proliferative actions observed in the normal mouse, the effects of GLP-2 are largely antiapoptotic in the neonatal pig on total parenteral nutrition (5) and are anti-inflammatory in rat models of intestinal disease (39). Nonetheless, numerous animal studies have shown that GLP-2 is beneficial in settings of intestinal dysfunction, including total parenteral nutrition (TPN)-induced atrophy, short bowel syndrome (SBS), inflammatory bowel disease (IBD), neonatal intestinal dysfunction, and gut injury caused by a variety of factors. Similarly, clinical trials have reported promising results for patients with SBS by increasing the capacity for enteral nutrient absorption. Trials for the use of GLP-2 in Crohn's disease are ongoing.
A Curious Receptor for GLP-2
The GLP-2 receptor (GLP-2R) is a G protein-coupled receptor (GPCR) belonging to the class B glucagon-secretin receptor family (29). The GLP-2R exhibits high homology with the receptors for glucagon, GLP-1, and glucose-dependent insulinotropic peptide, with which it shares some downstream signaling mediators, such as cAMP, protein kinase A (PKA), cAMP response element-binding protein (CREB), and AP-1 (46). The GLP-2R can initiate antiapoptotic responses through the cAMP pathway involving both PKA-dependent and -independent mechanisms (44). This antiapoptotic effect appears to involve inhibition of glycogen synthase kinase-3 (GSK-3) and Bad. Furthermore, in some cells, the GLP-2R can also couple to alternate G proteins (G
s and Gi/o) and can activate mitogen-activated protein kinase (MAPK) pathways (14, 24). A distinct feature of this receptor is the mechanisms by which it undergoes desensitization. Like many GPCRs, signaling through the GLP-2R is subject to both homologous and heterologous desensitization; however, in contrast to most known GPCRs, GLP-2R desensitization and internalization are independent of -arrestin and clathrin-mediated endocytosis (14, 15).
A major obstacle in the study of the GLP-2R has been the lack of cell lines that demonstrate endogenous GLP-2R signaling. It is therefore notable that the majority of studies addressing the mechanisms of GLP-2R signaling have utilized cell lines transfected with the GLP-2R, including 293-EBNA (29), COS (17, 29), BHK (14, 15, 25, 38, 44, 46), fibroblasts, DLD-1 colon cancer cells (14, 15), and HeLa cervical cancer cells (14, 24). Only a few cell lines have been identified that endogenously express the GLP-2R, including HeLa cells (24), CCD-18Co intestinal myofibroblasts (36) (and PED and PLB; unpublished data), and FHC fetal colon cells (36). Furthermore, in the case of HeLa (24) or CCD-18Co cells (and PED and PLB; unpublished data), the endogenous GLP-2R shows limited activity, with severely blunted cAMP accumulation or slight activation of MAPK, respectively. This paucity of robust cell models is relevant, as there may be significant differences between the signaling activities of the transfected and endogenous GLP-2R. For instance, the endogenous GLP-2R from hypothalamus (21), pituitary (21), intestinal mucosa (41), intestinal muscularis (38), and fetal intestine (10) produces a reproducible GLP-2 dose-response curve, in which peak cAMP accumulation occurs at moderate concentrations (10–10 to 10–9 M) but is reduced with higher levels of GLP-2, reminiscent of an "inverted U-shaped" dose-response. The mechanistic explanations for such a phenomenon are unknown but may potentially result from the unique nature of GLP-2R desensitization and trafficking (14, 15) or from dose-dependent coupling to alternate G proteins (24). Nevertheless, this characteristic dose-response is not observed in cells transfected with the GLP-2R, which demonstrate a maximum "plateau" of cAMP accumulation and only a slight reduction at concentrations of 10–6 M or greater, suggesting that these cells may not appropriately reflect the true nature of GLP-2R signaling in situ. Therefore, one major issue in the field is to determine which signaling mechanisms are relevant to the endogenous GLP-2R, especially as they relate to the intestinal growth effects of GLP-2, and to determine the intracellular pathways underlying the biological actions of GLP-2 with the use of relevant cell types.
While the signaling mechanisms of the GLP-2R remain unclear, its localization is equally curious and raises important questions regarding the mechanisms underlying GLP-2-induced intestinal growth. GLP-2R expression is restricted to the gastrointestinal tract and central nervous system, with limited expression in lung, cervix, and vagal afferents (24, 25, 29, 30, 45). However, the exact cellular localization of the GLP-2R, particularly in the intestine, has been a source of some controversy, with variable reports demonstrating expression in diverse endocrine cells, enteric neurons, and/or subepithelial myofibroblasts (SEMFs) (see Table 1). Some authors have suggested that these differences may be a result of methodological problems or species-specific expression; however, as each cell type has now been confirmed in several species by using multiple approaches, neither of these reasons seems likely, and it is possible that the GLP-2R is expressed in each of these diverse cellular localizations. While an explanation for the discrepancies between the studies remains elusive, it is nonetheless evident that neither crypt epithelial cells nor enterocytes express the GLP-2R. This has led to the hypothesis that GLP-2 requires an indirect signal, perhaps functioning through a paracrine mechanism, to induce its effects on intestinal growth (45).
|
Although there is a wealth of information about the ultimate effects of GLP-2 on intestinal physiology and on the signaling mechanisms initiated by the GLP-2R, there are relatively few studies addressing how these two events are functionally connected. Unraveling the mechanisms behind this has been a difficult task, which is understandable given the multiple interrelated actions of GLP-2 and the complex nature of the GLP-2R. It is clear that the diverse effects of GLP-2 require an indirect mechanism, likely involving not one but multiple indirect signals, interacting in a complementary fashion to effect different intestinal responses (i.e., proliferation, apoptosis, digestion, absorption, barrier function, blood flow, motility, anti-inflammation). Several studies have now provided important mechanistic data illustrating several of the indirect pathways of GLP-2 action (Fig. 1).
|
-cell insulin expression and secretion. However, in contrast to GLP-1 action in the -cell, GLP-2 does not appear to invoke intracellular calcium responses in primary SEMFs (PED and PLB; unpublished data) or in intestinal mucosal cells (41). Nevertheless, the regulation of IGF-I expression and/or secretion by cAMP-activating GPCR ligands is a mechanism common to several cell types, including glucagon in hepatocytes (23). In light of the emerging role of IGF-I as a mediator for GLP-2-induced growth, further study is clearly warranted to determine how and in which cell types GLP-2R signaling regulates IGF-I in the intestine.
IGF-I has been identified as a major mediator through which GLP-2 increases intestinal growth, specifically by inducing small intestinal crypt cell proliferation. This conclusion is supported primarily by studies of IGF-I knockout mice, which demonstrate marked impairments in small and large intestinal growth in response to GLP-2, compared with wild-type littermates, despite normal responses to other intestinal growth factors (10). Indeed, the effect of GLP-2 on crypt cell proliferation is completely lost in the absence of IGF-I. In normal mice, GLP-2 increases the rate of proliferation in the upper half of the intestinal crypts, thus expanding the number of cells responsible for populating the epithelium. This occurs in concert with a potential effect on the epithelial stem cells, as GLP-2 also increases the number of cells expressing musashi-1, a putative stem cell marker (10). The regulation of crypt cell proliferation involves several key mediators, most notably the canonical wingless (Wnt)/-catenin signaling system. The activation of -catenin transcriptional signaling by Wnt proteins, R-Spondin1, or IGF-I occurs through prevention of constitutive -catenin degradation, thereby allowing its translocation to the nucleus (8, 33). IGF-I appears to initiate -catenin signaling through a phosphatidylinositol-3 kinase (PI3K)-dependent pathway, involving Ras activation and GSK-3 inhibition (8). Conversely, inhibition of PI3K signaling through bone morphogenic protein (BMP) and phosphatase and tensin homolog (PTEN) downregulates -catenin signaling in the crypt cell (19). One role of -catenin in the intestine is to drive the transcription of genes required for proliferation, such as c-myc (28) while also inhibiting genes involved in terminal cell differentiation (33). This maintains cells within the active cell cycle, thereby increasing the overall numbers of proliferating cells. We have recently demonstrated that GLP-2 is a novel activator of -catenin signaling in the intestinal crypt, through a mechanism requiring IGF-I signaling through the IGF-IR (11). This provides a mechanistic basis for IGF-I as a mediator for GLP-2-induced proliferation and serves to link the GLP-2-IGF-I signaling system to other regulators of intestinal crypt cell fate, through common effects on -catenin (Fig. 2).
|
The essential role of IGF-I in GLP-2-induced crypt cell proliferation is contrasted by the limited impact of the loss of IGF-II, in which the enhancement of intestinal growth by GLP-2 is only partially blunted (10). IGF-II signals through several receptors, including the IGF-IR and isoform A of the insulin receptor, offering a potential explanation as to why the effects of IGF-II and IGF-I might differ. Indeed, the role of endogenous IGF-II in the regulation of intestinal growth is unique; IGF-II initiates a process known as crypt fission, in which whole crypts divide, resulting in an increase in the overall number of crypts (1). These additional crypts therefore become available to respond to other growth factors. This raises the possibility that, when faced with the absence of IGF-II, the intestine is unable to fully respond to GLP-2 due to the resulting impairment in crypt fission. This hypothesis would represent yet another level of complexity in the actions of GLP-2.
In addition to IGF-I, keratinocyte growth factor (KGF) has also been implicated in GLP-2-induced colonic growth. KGF is a member of the fibroblast growth factor family, expressed in SEMFs throughout the gastrointestinal tract, and is a tropic factor for epithelial cells (34). KGF colocalizes with the GLP-2R in SEMFs, and immunoneutralization of KGF in mice prevents the effect of GLP-2 on colonic weight and mucosal area without affecting colonic crypt depth or small intestinal growth (32). This is in contrast to the more marked effect of IGF-I ablation on both small and large intestinal growth responses to GLP-2. Moreover, unlike IGF-I, KGF treatment produces differential effects to those of GLP-2, affecting mainly colonic growth and the differentiation of goblet cells (2, 22, 42). Indeed, Ørskov et al. (32) found that the GLP-2-induced increase in mucin expression was blocked by KGF antibodies, suggesting that the specific role of KGF may be to promote colonic goblet cell differentiation in response to GLP-2. Nonetheless, it is possible that immunoneutralization may be insufficient to block small intestinal KGF, and therefore additional studies, utilizing a knockout model, would be helpful to determine whether this effect of KGF is truly restricted to the colon. Furthermore, a more detailed study of the role for KGF in crypt cell proliferative responses would help to determine whether the KGF effect extends beyond that on the goblet cells and whether KGF may interact with the proliferative actions of IGF-I.
These studies have only begun to uncover the mechanisms through which GLP-2 induces growth in the normal intestine. Although proliferative responses in the small intestinal crypt appear to require IGF-I, it is unknown how GLP-2 alters apoptosis, permeability, or nutrient digestion or absorption in the villus epithelium. Cheeseman (6) reported that GLP-2-mediated epithelial glucose uptake, through the sodium-dependent glucose transporter, occurred through a PI3K-dependent mechanism. Furthermore, the activation of Akt in the intestinal mucosa has been implicated in the antiapoptotic actions of GLP-2 (3, 9). The PI3K-Akt pathway is a common signaling effector for multiple growth factors and cytokines, including IGF-I. However, although IGF-I activates Akt signaling in the intestinal mucosa, it appears that, unlike the response observed with -catenin in the crypt, IGF-I signaling is not strictly required for the activation of Akt by GLP-2 (11). This suggests that, in addition to the requirement of IGF-I for proliferation, other factors exist that mediate the diverse biological actions of GLP-2. The SEMFs, for example, express and secrete a wide range of different growth factors and cytokines in addition to the IGFs and KGF that may participate in some of the effects of GLP-2 (34).
Several studies have uncovered an exciting role for GLP-2R signaling in submucosal enteric neurons. Functioning through a nitric oxide (NO)-dependent mechanism, GLP-2 acutely and dose-dependently increases intestinal blood flow (17, 18). This is associated with the activation of endothelial nitric oxide synthase (eNOS), which may be a direct consequence of GLP-2R signaling, as PKA-mediated phosphorylation at Ser1177 is sufficient to increase eNOS activity (43). However, as Ser1177 is also a phosphorylation site for Akt, this activation may be indirect, through another paracrine factor. This is potentially relevant, as Akt is a downstream kinase in IGF-IR signaling and is activated by GLP-2 treatment in the intestinal mucosa in an acute and sustained manner (3, 9). Very recently, Sigalet et al. (39) have reported that GLP-2 reduced intestinal damage and the levels of inflammatory cytokines in a rat model of IBD through a mechanism requiring vasoactive intestinal polypeptide (VIP)-expressing submucosal enteric neurons. One of the curious findings of this report was that, in a counterintuitive manner, GLP-2 administration reduced the rate of epithelial proliferation in this setting; given that inflammation itself induces crypt cell proliferation (27), it would seem that this effect of GLP-2 may be an indirect result of a VIP-dependent anti-inflammatory action (39). It is thus possible that NO or VIP signaling may affect GLP-2-induced intestinal growth through several indirect mechanisms. For instance, NO-induced alterations in local circulation may function as a permissive mechanism for intestinal growth; indeed, a reduction in blood flow is associated with intestinal atrophy (31). However, a direct effect of NO or VIP on epithelial growth cannot be ruled out (12, 37). Finally, although a direct growth effect of neuronal GLP-2R signaling is unknown, Bjerknes and Cheng (2) have shown that GLP-2-induced c-fos expression in the intestinal crypt is dependent on a tetrodotoxin-sensitive mechanism; a neural mechanism in the control of crypt epithelial function therefore cannot be ruled out.
Finally, the spectrum of GLP-2 actions in the intestine suggests that GLP-2-based therapy may be promising in multiple clinical settings, including TPN, SBS, neonatal intestinal dysfunction, IBD, and intestinal injury. One potential benefit is that GLP-2 therapy may be associated with fewer extragastrointestinal sequelae compared with other growth factors, given the relatively intestine-specific effects of GLP-2 in animal models. However, in light of the indirect mechanisms of GLP-2 action (Fig. 1), the mediators involved deserve consideration when the benefits and potential side effects of GLP-2 administration are being determined. For instance, it has been proposed that GLP-2 may engage alternate mediators in a dose-dependent fashion (i.e., proliferative vs. anti-inflammatory responses) (39), suggesting differential clinical benefits depending on the dose of GLP-2, as well as on the condition being treated. Furthermore, given the proliferative effects of GLP-2, one potential detriment to prolonged GLP-2 therapy may be an increased risk for tumorigenesis (7, 16). Indeed, at least one study has shown that GLP-2 administration in mice can accelerate the growth of chemically induced intestinal tumors (40). This may be especially important in IBD patients, who already possess an increased risk of gastrointestinal cancer due to chronic inflammation. Therefore, careful dose regulation and monitoring for gastrointestinal dysplasia should be considered in any clinical use of GLP-2.
Summary
It has become clear that GLP-2 functions through multiple interrelated pathways that defy simple definition. Each new discovery in this field raises important new questions, both for the actions of GLP-2 and for the integrated physiology of the intestine. Future studies will have to consider the full range of GLP-2 effects, as well as all the sites of GLP-2R expression, to elucidate the multiple mechanisms of GLP-2 action. In particular, the function of the GLP-2R in intestinal endocrine cells is unknown, although it is tempting to hypothesize a role in the regulation of intestinal peptide hormone and/or serotonin secretion. Furthermore, the recent description of GLP-2R expression in vagal afferents (30), as well as its widespread expression in the central nervous system, hypothalamus, and pituitary (25), suggest that the intestinal functions of GLP-2 may be functionally linked to currently unidentified whole body actions. Indeed, recent studies have identified bone as a target of GLP-2 action, although how these effects are mediated is completely unknown (20). A major goal of future research should therefore be not only to discover each mediator but also to determine how each interacts and contributes functionally to the diverse actions of GLP-2.
GRANTS
Studies on GLP-2 in the laboratory are supported by an operating grant from the Canadian Institutes of Health Research (CIHR). P. E. Dubé was supported by a Doctoral Research Award from the CIHR in partnership with the Canadian Digestive Health Foundation, and P. L. Brubaker by the Canada Research Chairs Program.
FOOTNOTES
Address for reprint requests and other correspondence: P. L. Brubaker, Rm. 3366, 1 King's College Cir., Toronto, ON M5S 1A8, Canada (e-mail: p.brubaker{at}utoronto.ca)
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
-catenin signaling in mouse intestinal crypt through the insulin-like growth factor-1 pathway. Gastroenterology 132: S, 2007.This article has been cited by other articles:
|
|
 |
|
  X. Liu, S. G. Murali, J. J. Holst, and D. M. Ney Whey protein potentiates the intestinotrophic action of glucagon-like peptide-2 in parenterally fed rats Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2009; 297(5): R1554 - R1562. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Baldassano, S. Liu, M.-H. Qu, F. Mule, and J. D. Wood Glucagon-like peptide-2 modulates neurally evoked mucosal chloride secretion in guinea pig small intestine in vitro Am J Physiol Gastrointest Liver Physiol, October 1, 2009; 297(4): G800 - G805. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P D Cani, S Possemiers, T Van de Wiele, Y Guiot, A Everard, O Rottier, L Geurts, D Naslain, A Neyrinck, D M Lambert, et al. Changes in gut microbiota control inflammation in obese mice through a mechanism involving GLP-2-driven improvement of gut permeability Gut, August 1, 2009; 58(8): 1091 - 1103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Izumi, S. Ishizuka, A. Inafune, T. Hira, K. Ozawa, T. Shimizu, M. Takase, and H. Hara {alpha}-Lactalbumin Hydrolysate Stimulates Glucagon-Like Peptide-2 Secretion and Small Intestinal Growth in Suckling Rats J. Nutr., July 1, 2009; 139(7): 1322 - 1327. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Amato, S. Baldassano, R. Serio, and F. Mule Glucagon-like peptide-2 relaxes mouse stomach through vasoactive intestinal peptide release Am J Physiol Gastrointest Liver Physiol, March 1, 2009; 296(3): G678 - G684. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. Garrison, C. M. Dekaney, D. C. von Allmen, P. K. Lund, S. J. Henning, and M. A. Helmrath Early but not late administration of glucagon-like peptide-2 following ileo-cecal resection augments putative intestinal stem cell expansion Am J Physiol Gastrointest Liver Physiol, March 1, 2009; 296(3): G643 - G650. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Liu, S. G. Murali, J. J. Holst, and D. M. Ney Enteral nutrients potentiate the intestinotrophic action of glucagon-like peptide-2 in association with increased insulin-like growth factor-I responses in rats Am J Physiol Regulatory Integrative Comp Physiol, December 1, 2008; 295(6): R1794 - R1802. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
