`Copyright c(cid:176) 2002 Taylor & Francis
`1060-6823/02 $12.00 + .00
`DOI: 10.1080/10606820290005155
`
`Structure-Function of the Glucagon Receptor Family
`of G Protein–Coupled Receptors: The Glucagon, GIP,
`GLP-1, and GLP-2 Receptors
`
`P. L. Brubaker
`Departments of Physiology and Medicine, University of Toronto, Toronto, Ontario, Canada
`D. J. Drucker
`Department of Medicine, Banting and Best Diabetes Centre, Toronto General Hospital,
`Toronto, Ontario, Canada
`
`The glucagon-like peptides include glucagon, GLP-1, and
`GLP-2, and exert diverse actions on nutrient intake, gastrointesti-
`nal motility, islet hormone secretion, cell proliferation and apopto-
`sis, nutrient absorption, and nutrient assimilation. GIP, a related
`member of the glucagon peptide superfamily, also regulates nutri-
`ent disposal via stimulation of insulin secretion. The actions of these
`peptides are mediated by distinct members of the glucagon recep-
`tor superfamily of G protein–coupled receptors. These receptors
`exhibit unique patterns of tissue-specific expression, exhibit consid-
`erable amino acid sequence identity, and share similar structural
`and functional properties with respect to ligand binding and sig-
`nal transduction. This article provides an overview of the biology
`of these receptors with an emphasis on understanding the unique
`actions of glucagon-related peptides through studies of the biology
`of their cognate receptors.
`
`Keywords Diabetes, GIP, GLP-1, GLP-2, Glucagon, Glucose,
`Intestine, Regulatory Peptides
`
`GENERAL INTRODUCTION
`The gastroenteropancreatic-brain axis expresses a diverse
`number of peptide hormones, including several that are struc-
`turally related to the pancreatic hormone, glucagon (Table 1).
`Within the mammalian glucagon superfamily of peptides,
`glucagon and the glucagon-like peptides, GLP-1 and GLP-2,
`are all encoded by a single proglucagon gene (Bell et al. 1983;
`Irwin 2001; White and Saunders 1986). Tissue-specific
`posttranslational processing of proglucagon by prohormone
`
`Address correspondence to Daniel J. Drucker, M.D., Banting and
`Best Diabetes Centre, Toronto General Hospital, 200 Elizabeth Street,
`Toronto, Ontario, M5G 2C4 Canada. E-mail: d.drucker@utoronto.ca
`
`convertases results in the liberation of glucagon in the pancreatic
`A cell, and GLP-1 and GLP-2 in the intestinal L cell and brain
`(Mojsov et al. 1986; Orskov et al. 1987). As discussed below,
`all three proglucagon-derived peptides (PGDPs) play impor-
`tant roles in the physiologic regulation of nutrient homeosta-
`sis, through effects on energy intake and satiety, nutrient fluxes
`through and across the gastrointestinal tract, and energy as-
`similation. Several of these biological activities are shared by
`a fourth glucagon-related peptide hormone, glucose-dependent
`insulinotropic peptide (GIP) (Table 1). The diverse biological
`activities of these glucagon-related peptides are exerted through
`highly specific G protein–coupled receptors (GPCR), members
`of the glucagon receptor superfamily. For the purposes of this
`article, the biological activities and mechanisms of action of
`each of these peptides will be discussed individually.
`
`Biological Actions of the Glucagon-Related Peptides
`Glucagon
`Glucagon is a 29 amino acid peptide hormone liberated from
`islet A cells in the endocrine pancreas. The principal actions
`of glucagon involve regulation of metabolic pathways involved
`in glucose homeostasis. Glucagon secretion from the A cell is
`coupled to the ambient levels of circulating glucose, with hypo-
`glycemia and hyperglycemia leading to stimulation and inhibi-
`tion of glucagon release, respectively (Unger 1985). Glucagon
`action in the liver impinges on numerous enzymes important
`for control of gluconeogenesis, glycogenolysis, and fatty acid
`metabolism, leading to restoration of circulating glucose con-
`centrations in the setting of energy depletion. The actions of
`glucagon on hepatic glucose production are exerted at multiple
`levels, including regulation of gene transcription and modula-
`tion of enzyme activity (Pilkis and Granner 1992). Defective
`
`179
`
`Page 1
`
`NPS EX. 2038
`CFAD v. NPS
`IPR2015-00990
`
`
`
`180
`
`P. L. BRUBAKER AND D. J. DRUCKER
`
`TABLE 1
`The amino acid sequences of human glucagon, GLP-1, GLP-2, and GIP
`
`Glucagon
`GLP-1
`GLP-2
`GIP
`
`HSQGTFTSDYSKYLDSRRAQDFVQWLMNT
`HAEGTFTSDVSSYLEGQAAKEFIAWLVKGRG
`HADGSFSDEMNTILDNLAARDFINWLIQTKITD
`YAEGTFISDYSIAMDKIRQQDFVNWLLAQ
`
`glucagon secretion contributes to the development of hypo-
`glycemia in insulin-treated patients with diabetes (Gerich 1988;
`Taborsky, Jr. et al. 1998).
`Glucagon receptors are widely expressed in multiple tissues
`including the liver, brain, pancreas, heart, kidney, and smooth
`muscle cell in the gastrointestinal tract and peripheral vascula-
`ture. In the heart, glucagon stimulates inotropic and chronotropic
`activity (Parmley et al. 1968), leading to the selective use of
`glucagon in medical emergencies characterized by refractory
`bradycardia due to beta blocker toxicity or cardiogenic shock
`(White 1999). Supraphysiological levels of glucagon also exert
`vasodilator effects by reducing vascular resistance in specific
`vascular beds (Farah 1983).
`Glucagon infusion activates nephrogenic adenylyl cyclase
`(leading to increased urinary cAMP excretion), increases the
`glomerular filtration rate, and regulates ion transport and elec-
`trolyte excretion in the kidney (Ahloulay et al. 1992; Broadus
`et al. 1970; Schwartz Sorensen et al. 1993). Although hypo-
`glycemia is associated with increased hepatic and renal glu-
`cose production, a role for glucagon in the regulation of renal
`gluconeogenesis is unclear (Cersosimo et al. 1999; Stumvoll
`et al. 1997, 1998). Whether glucagon actions on the kidney
`are essential for regulation of renal physiology remains to be
`determined.
`Within the brain, PGDPs, including glucagon, are synthe-
`sized principally in the brainstem and, to a lesser extent, in
`the hypothalamus (Drucker and Asa 1988); however, glucagon-
`immunoreactive nerve fibers are widely distributed through the
`mammalian central nervous system (Dorn et al. 1983). Glucagon
`administration elicits a broad spectrum of actions, including
`stimulation of anterior pituitary hormone secretion (Spathis et al.
`1974), hypothalamic somatostatin release (Shimatsu et al. 1983),
`and ketone utilization (Kirsch and D’Alecy 1984). Intracere-
`broventricular administration of glucagon potently suppresses
`food intake in the rat (Inokuchi et al. 1984) and enhances sym-
`pathetic nerve activity (Krzeski et al. 1989; Shimizu et al. 1993),
`whereas glucagon suppresses the activity of hypothalamic
`glucose-sensitive neurons but has no effect on cortical neurons
`in the rat (Inokuchi et al. 1986).
`Glucagon stimulates lipolysis in isolated adipocyte prepara-
`tions from young animals (Heckemeyer et al. 1983; Lefebvre and
`Luyckx 1969); however, these actions are significantly attenu-
`ated with aging or weight gain (Bertrand et al. 1980). The poten-
`tial importance of glucagon for stimulation of human adipocyte
`lipolysis is less clear (Richter et al. 1989; Richter and Schwandt
`1985). Glucagon withdrawal or physiological hyperglucagone-
`
`mia did not produce significant changes in palmitate flux in nor-
`mal or diabetic human subjects (Jensen et al. 1991). Similarly,
`subcutaneous infusion of glucagon into the gastrocnemius mus-
`cle or abdominal tissue had no effects on lipolysis rates or blood
`flow in normal human subjects (Bertin et al. 2001; Gravholt
`et al. 2001). Glucagon modulates electrical activity in gastroin-
`testinal smooth muscle (Patel et al. 1979; Taylor et al. 1975),
`leading to its clinical use as spasmolytic agent for modulation
`of motility during endoscopic and radiologic examination of the
`gastrointestinal tract.
`
`Glucagon-Like Peptide-1 and Glucose-Dependent
`Insulinotropic Peptide
`The incretin concept was first developed in 1930 to describe
`the observation that administration of glucose by the oral route
`was associated with a greater increment in insulin secretion than
`was a euglycemic load given intravenously (Elrick et al. 1964;
`LaBarre and Still 1930). This concept was further refined by both
`Unger (Unger and Eisentraut 1969) and Creutzfeldt (Creutzfeldt
`1979) to define an incretin as a gut hormone released in response
`to nutrient ingestion that stimulates glucose-dependent insulin
`secretion. The first incretin to be identified was the duodenal
`hormone, GIP. Although initially reported to possess gastric in-
`hibitory actions (Brown 1982; Brown and Dryburgh 1971), GIP
`was subsequently discovered to be a potent stimulator of insulin
`secretion (Dupre et al. 1973). Subsequent studies using GIP re-
`ceptor antagonists have clearly established that GIP fulfills all
`of the requirements for an incretin (Baggio et al. 2000; Lewis
`et al. 2000; Tseng, Kieffer et al. 1996). However, reports that
`immunoneutralization of GIP does not abolish the incretin effect
`of ingested nutrients (Ebert and Creutzfeldt 1982; Ebert et al.
`1983), led to an ongoing search for additional insulinotropic
`gut hormones. It was not until 1987 that the second major in-
`cretin was discovered to be another member of the glucagon-
`family of peptides, GLP-1 (Kreymann et al. 1987; Mojsov et al.
`1987; Orskov and Nielsen 1988). Numerous studies on the in-
`sulinotropic actions of GIP and GLP-1 have now established
`that, together, these peptides constitute the majority of the in-
`cretin effect following ingestion of a meal, and that the two hor-
`mones contribute equivalently to this effect (Kieffer and Habener
`1999). The importance of GIP and GLP-1 in glucose homeosta-
`sis is suggested by the strong degree of sequence conservation
`that exists for each of these hormones (Irwin 2001). Consistent
`with this hypothesis, receptor antagonist studies and the analysis
`of mice with null mutations in the receptors for GIP (GIPR) or
`
`Page 2
`
`
`
`STRUCTURE-FUNCTION OF G PROTEIN–COUPLED RECEPTORS
`
`181
`
`GLP-1 (GLP-1R) have demonstrated that loss of GIP or GLP-
`1 action results in an impaired insulin response to oral glucose
`(Miyawaki et al. 1999; Scrocchi et al. 1996). Whether these pep-
`tides are the sole hormones mediating the incretin effect awaits
`the development of mice with combined defects in both GIP
`and GLP-1 receptor signaling. Finally, very recent studies have
`indicated an exciting role for GLP-1 as a fl cell tropic factor.
`Administration of GLP-1, or exendin-4, a long-acting GLP-1R
`agonist, to mice or rats with diminished fl cell reserve leads to
`enhancements in fl cell mass and an increased capacity for in-
`sulin secretion (Stoffers et al. 2000; Xu et al. 1999). Consistent
`with these observations, mice with a null mutation in the GLP-
`1R exhibit alterations in fl cell topography and reduced insulin
`secretion (Ling et al. 2001). Whether GIP also regulates fl cell
`growth remains to be established.
`Because of their insulinotropic and glucose-lowering actions,
`both GIP and GLP-1 are candidate peptides for the treatment
`of type 2 diabetes mellitus. However, only GLP-1 is currently
`in clinical trials for such therapeutic use, for two main rea-
`sons. First, although both GIP and GLP-1 inhibit gastric emp-
`tying (Nauck et al. 1996), an effect that delays postprandial
`rises in glycemia, only GLP-1 also induces satiety (Scrocchi
`et al. 1996; Turton et al. 1996). In addition, these peptides
`have opposing effects on glucagon release, with GLP-1 inhibit-
`ing, but GIP stimulating, glucagon secretion (Elahi et al. 1979;
`Komatsu et al. 1989). More importantly, although GLP-1 low-
`ers glycemia in patients with type 2 diabetes (Gutniak et al.
`1992; Nauck et al. 1996), the insulinotropic actions of GIP ap-
`pear to be markedly abrogated in such individuals (Elahi et al.
`1994; Nauck et al. 1993). These findings are consistent with
`previous observations that the incretin effect is diminished in
`patients with type 2 diabetes (Nauck et al. 1986), but, in addi-
`tion, suggest that this defect occurs consequent to abnormalities
`at the level of the GIPR and/or in its postreceptor signaling cas-
`cade. Indeed, diabetic Zucker fatty rats, a rodent model of type 2
`diabetes, exhibit decreased islet expression of the GIPR in as-
`sociation with diminished fl cell responsiveness to GIP (Lynn
`et al. 2001).
`
`Glucagon-Like Peptide-2
`GLP-2, located carboxyterminal to GLP-1 in the proglucagon
`sequence, is cosecreted with GLP-1, oxyntomodulin, and gli-
`centin from gut L cells, primarily in response to nutrient in-
`gestion (Brubaker, Crivici et al. 1997; Xiao et al. 1999; Xiao,
`Boushey et al. 2000). Following clinical and experimental ob-
`servations linking intestinal growth, injury, and gut adaptation
`to increased levels of circulating PGDPs (Drucker 1999, 2001),
`GLP-2 was subsequently identified as the PGDP exhibiting sig-
`nificant intestinotrophic properties in vivo (Drucker et al. 1996).
`Administration of GLP-2 to normal mice and rats results in
`significant increases in mucosal thickness of the small and large
`intestine, with more prominent trophic effects consistently ob-
`served in the small bowel, especially in the jejunum (Drucker et al.
`1996; Drucker, Deforest et al. 1997; Drucker, Shi et al. 1997;
`
`Tsai, Hill, Asa et al. 1997; Tsai, Hill, and Drucker 1997). The
`intestinotrophic effects of GLP-2 are mediated by stimulation
`of crypt cell proliferation and inhibition of apoptosis in both
`the crypt and villus compartments of the small bowel (Drucker
`et al. 1996; Tsai, Hill, Asa et al. 1997). GLP-2 administration
`also results in inhibition of gastric emptying, reduced gastric
`acid secretion, stimulation of intestinal hexose transport and
`nutrient absorption, and reduction in mucosal epithelial perme-
`ability in the small bowel (Benjamin et al. 2000; Brubaker, Izzo
`et al. 1997; Cheeseman and Tsang 1996; Wojdemann et al. 1998,
`1999).
`The reparative and antiapoptotic actions of GLP-2 have also
`been examined in the setting of experimental intestinal injury.
`GLP-2 reduces mucosal damage and decreases morbidity and
`mortality in a broad spectrum of rodent models of small and
`large bowel injury, including enteral nutrient deprivation, intesti-
`nal inflammation, ischemia, and chemotherapy-induced enteritis
`(Alavi et al. 2000; Boushey et al. 1999, 2001; Chance et al. 1997,
`2001; Drucker et al. 1999). GLP-2 also enhances nutrient ab-
`sorption and intestinal adaptation in rats with experimental small
`bowel resection (Scott et al. 1998) and in human subjects with
`short bowel syndrome (Jeppesen et al. 2001).
`
`RECEPTOR CLONING: SPECIFICITY
`AND DISTRIBUTION
`The GLP-1R was the first member of this group of receptors to
`be cloned (Thorens 1992). A 463 amino acid 7 transmembrane-
`spanning protein, the GLP-1R exhibits 27% to 40% sequence
`homology to the receptors for secretin, calcitonin, and parathy-
`roid hormone. As these receptors shared higher identity with
`each other than with other members of the G protein–coupled
`receptor superfamily, they were classified together into a new
`family of receptors (Thorens 1992) now known as the type II
`receptor family. However, despite the strong sequence homol-
`ogy between GLP-1 and the other members of the glucagon-
`related family of peptides, the GLP-1R recognizes GLP-1 specif-
`ically, with no demonstrable binding by a number of related
`peptides, including secretin and vasoactive intestinal peptide
`(VIP) (Fehmann et al. 1994). Although binding of GIP and
`glucagon to the GLP-1R has been observed in some studies, this
`appears to occur only at micromolar concentrations (Thorens
`1992). Furthermore, only GLP-1 appears to activate the GLP-
`1R, stimulating production of cAMP with an ED50 of 3 nM in
`COS cells transfected with the human receptor (Dillon et al.
`1993). Consistent with the very high degree of sequence conser-
`vation found for GLP-1 in most mammalian species, the GLP-
`1R also exhibits conservation between species, with 90% ho-
`mology between the rat and human receptors (Sivarajah et al.
`2001). The distribution of the mammalian GLP-1 receptor also
`appears to be conserved, with RT-PCR and Northern blot analy-
`ses demonstrating expression in tissues known to be responsive
`to GLP-1, including the pancreatic fl cells, the stomach, and
`the brain (Bullock et al. 1996; Campos et al. 1994; Thorens
`1992).
`
`Page 3
`
`
`
`182
`
`P. L. BRUBAKER AND D. J. DRUCKER
`
`The rat glucagon receptor (GR) cDNA was cloned in 1993
`and encodes a 485 amino acid protein of »54,962 daltons
`(Jelinek et al. 1993). Glucagon stimulates an increase in cAMP
`and intracellular calcium in transfected cells expressing the
`cloned receptor, with much weaker receptor activation noted
`only with pharmacological concentrations of the related pep-
`tides secretin, calcitonin, and parathyroid hormone (Jelinek et al.
`1993). The human glucagon receptor cDNA encodes a protein
`of 477 amino acids that exhibits 82% identity to the rat receptor
`(Lok et al. 1994; MacNeil et al. 1994). Glucagon inhibits binding
`of [125-I]-glucagon to the cloned human receptor with an IC50
`of 5 nM, with tenfold lower tracer displacement exhibited by
`oxyntomodulin and GLP-1, whereas GIP, GLP-2, and secretin
`fail to displace binding of [125-I]-glucagon binding at concen-
`trations up to 3 uM (MacNeil et al. 1994). Mammalian glucagon
`receptor RNA transcripts are widely expressed in peripheral tis-
`sues, including liver, kidney, heart, adipose tissue, pancreatic
`islets, the gastrointestinal tract, spleen, thymus, adrenal gland,
`ovary, testes, and the central nervous system (Campos et al.
`1994; Dunphy et al. 1998; Hansen et al. 1995; Svoboda et al.
`1994).
`Cloning of the GIPR in 1993 demonstrated that this 455
`amino acid protein also belongs to the 7 transmembrane-
`spanning, glucagon receptor-related superfamily of receptors,
`sharing 44% sequence identity with the glucagon receptor (Usdin
`et al. 1993). The GIPR exhibits highly specific GIP binding,
`although some binding by exendin-4, a GLP-1R agonist, was
`noted at very high concentrations of 1–10 „M (Gremlich et al.
`1995). However, there was no demonstrable response of the
`GIPR to other members of the glucagon-related family of hor-
`mones (Gremlich et al. 1995; Usdin et al. 1993). A high degree
`of sequence conservation (>80%) between the rat, hamster, and
`human GIPR’s has been reported (Gremlich et al. 1995; Yamada
`et al. 1995). Consistent with its known sites of action, the GIP
`receptor is expressed in tissues such as the pancreas and upper
`gastrointestinal tract (Usdin et al. 1993). Interestingly, the GIPR
`is also expressed at low levels in the adrenal gland, where over-
`expression appears to contribute to the development of “Food-
`Dependent Cushing’s Syndrome” (N’Diaye et al. 1998). The
`receptor is also found in adipose tissue, where it may play a role
`in lipolysis (Yip and Wolfe 2000), and in the brain, where GIP
`function remains to be ascertained. GIP also exerts anabolic ac-
`tions in bone cells (Bollag et al. 2000), and GIP administration
`prevents ovariectomy-associated bone loss in vivo (Bollag et al.
`2001).
`The human and rat GLP-2 receptor cDNAs were isolated from
`intestinal and hypothalamic cDNA libraries in 1999 (Munroe
`et al. 1999). The rat GLP-2R cDNA comprises 2357 nucleotides
`and encodes a 550 amino acid receptor precursor protein. Two
`putative functionally identical translation start sites are present at
`0
`end of the rat receptor, whereas only a single upstream ATG
`the 5
`is present in the human sequence; the mouse GLP-2R sequence
`contains only the downstream ATG codon (Lovshin et al. 2001).
`The GLP-2R specifically recognizes GLP-2 with an EC50 for
`
`cAMP stimulation of 0.58 nM, with no significant stimulation of
`cAMP generation detected with related glucagon-like peptides
`and peptide members of the glucagon-secretin superfamily at
`concentrations of 10 nM (DaCambra et al. 2000; Munroe et al.
`1999). Unlike the more widely expressed GLP-1 and glucagon
`receptors, GLP-2R expression is more restricted, predominantly
`to endocrine cells and enteric neurons in the gastrointestinal
`tract and selected brain neurons (Bjerknes and Cheng 2001;
`Lovshin et al. 2001; Munroe et al. 1999; Yusta, Huang et al.
`2000). GLP-2R expression has been localized to distinct subsets
`of enteroendocrine cells in the human stomach and both small
`and large intestine, implying that GLP-2R receptor signaling in
`gut endocrine cells stimulates the activation and/or repression
`of endocrine-derived downstream mediators of GLP-2 action
`(Yusta, Huang et al. 2000).
`
`GENOMIC ORGANIZATION
`The human, mouse, and rat glucagon receptor genes contain
`13 exons, and the human gene has been mapped to chromo-
`some 17q24 (Burcelin et al. 1995; Lok et al. 1994; Maget et al.
`1994), with the relative location of introns conserved among
`members of the receptor superfamily. Although RT-PCR stud-
`ies have identified incompletely processed glucagon receptor
`transcripts, with the potential to give rise to variant glucagon re-
`ceptor proteins (Maget et al. 1994), variant receptor transcripts
`have not been detected in subsequent studies (Dunphy et al.
`1998; Hansen et al. 1995), hence the significance of potential
`glucagon receptor splice variants requires further clarification.
`The human GIPR gene contains 14 exons (Yamada et al.
`1995), whereas the rat gene contains 16 exons (Boylan et al.
`1999). The human GIPR gene is located on chromosome 19q13.3
`(Gremlich et al. 1995). As is the case with the glucagon, GLP-1,
`0
`and GLP-2 receptor genes (Geiger et al. 2000), the GIPR 5
`-
`flanking region contains several Sp1 binding sites, but no func-
`tional TATA or CAAT boxes (Boylan et al. 1999; Geiger et al.
`2000; Lovshin et al. 2001; Wildhage et al. 1999). The human
`GLP-1R gene is located on band 21 of chromosome 6 (Stoffel
`et al. 1993), whereas the GLP-2R gene has been localized to hu-
`man chromosome 17p13.3 (Munroe et al. 1999). The structural
`organization of the GLP-1 and GLP-2 receptor genes remains
`incompletely characterized.
`
`ALTERNATIVE RNA SPLICING
`RNA splicing resulting in functionally distinct receptor iso-
`forms has not been conclusively proven for any of the glucagon-
`receptor related receptors with the exception of the GIPR. Ge-
`nomic analysis of the GIPR identified one exon, designated
`#8a, that appears to be alternatively spliced, resulting in a vari-
`ant mRNA transcript that encodes a truncated receptor (Boylan
`et al. 1999). Although the functional properties of the variant
`receptor are unknown, an internally deleted GIPR, missing nu-
`cleotides 793-924, that does not bind GIP in transient expression
`assays, has been isolated (Gremlich et al. 1995). A second mutant
`
`Page 4
`
`
`
`STRUCTURE-FUNCTION OF G PROTEIN–COUPLED RECEPTORS
`
`183
`
`receptor cDNA, isolated from human insulinoma cells, lacks an
`internal 62 bp exon, resulting in a premature stop codon that
`encodes a truncated receptor RNA (Volz et al. 1995). Similar
`variant GIP receptor transcripts were detected in RNA from en-
`dothelial cell lines (Zhong et al. 2000). In contrast, a second
`human GIPR with a 27 amino acid insertion in the C-terminal
`tail appears to be fully functional (Gremlich et al. 1995); how-
`ever, the tissue distribution and functional significance of this
`receptor has not been examined in human tissues in vivo.
`
`SIGNALING/G PROTEIN INTERACTIONS
`Glucagon activates multiple G protein–mediated signal trans-
`duction pathways in liver cells, leading to stimulation of ade-
`nylyl cyclase (Pohl et al. 1969) and phosphoinositol turnover
`(Wakelam et al. 1986). The importance of dual signaling sys-
`tems is exemplified by studies with glucagon analogues that fail
`to stimulate cAMP formation, yet activate glycogenolysis and
`gluconeogenesis, via stimulation of glucagon receptor signal-
`ing, coupled to inositol phosphate production (Wakelam et al.
`1986). Although both glucagon and fl-adrenergic agonists stim-
`ulate adenylyl cyclase in liver cells, the glucagon receptor is
`coupled to both Gsfi-s and Gsfi-L whereas hepatic fl-adrenergic
`receptors are primarily coupled to Gsfi-L (Yagami 1995). Gener-
`ation of the (H178R) mutation in the glucagon receptor results in
`increased levels of basal cAMP and increased glucagon binding
`affinity (Hjorth et al. 1998). However, glucagon-mediated stimu-
`lation of cAMP or intracellular calcium appears to be dependent
`on amino acid sequences residing within the second and third
`intracellular loops (Cypess et al. 1999). Furthermore, deletional
`studies have demonstrated that the majority of carboxytermi-
`nal amino acid sequences distal to TM7 are not required for
`glucagon binding or adenylyl cyclase activation (Unson et al.
`1995).
`Differential interaction of the glucagon receptor with specific
`Gs isoforms has been detected in studies using GR-Gfis fusion
`proteins. The affinity of the wild-type GR for glucagon (»IC50
`of »46 nM) was comparable to the affinity observed with the
`GR-Gfis-S hybrid receptor, whereas the GR-Gfis-L receptor ex-
`hibited a lower IC50 for glucagon binding of »3.2 nM and a
`fourfold higher IC50 for adenylyl cyclase activation compared
`to GR-Gfis-L (Unson et al. 2000). The higher affinity reversed
`in the presence of GTP(cid:176) S, suggesting that specific coupling of
`the GR to distinct Gs isoforms may account for the dual class of
`high and low affinity glucagon receptors described in previous
`studies (Horwitz et al. 1985, 1986).
`Although coupling of the GLP-1R to the adenylyl cyclase
`pathway has been well established (Drucker et al. 1987), GLP-1
`also activates phospholipase C and MAPK in some cell types
`(Montrose-Rafizadeh et al. 1999; Wheeler et al. 1993), and stim-
`ulates calcium influx leading to increases in cytosolic free cal-
`cium [Ca2C
`]i. Consistent with this observation, the GLP-1R cou-
`ples to multiple G proteins, including Gfis, Gfiq=11 and Gfii1;2
`(Montrose-Rafizadeh et al. 1999) and a GLP-1R-G protein in-
`teraction domain includes amino acid sequences within the third
`
`intracellular loop (Hallbrink et al. 2001). Consistent with these
`findings, structural comparison of the GLP-1R, with related
`members of the GPCR superfamily, implicated the sequence
`K334-L335-K336 in the 3rd intracellular loop as a potential site
`for G protein coupling, and deletion of these amino acids abol-
`ishes both coupling to adenylyl cyclase and insulin secretion
`(Salapatek et al. 1999; Takhar et al. 1996). Alanine scanning
`mutagenesis further identified amino acids V327, I328, V331,
`and K334 as essential for the stimulation of adenylyl cyclase by
`GLP-1 (Mathi et al. 1997). Interestingly, 3-dimensional model-
`ing of the 3rd intracellular loop demonstrates that these amino
`acids form one face of an fi-helix, similar to that reported for the
`G protein coupling region of the m5 muscarinic receptor (Hill-
`Eubanks et al. 1996). The importance of amino acids within
`the third intracellular loop for G protein coupling is further
`emphasized by experiments using synthetic peptide sequences
`that demonstrate activation of both pertussis- and cholera toxin-
`sensitive G proteins as assessed by measurement of GTPase
`activity (Hallbrink et al. 2001).
`GIP also stimulates adenylyl cyclase activation in islet and
`intestinal cells and in heterologous cells expressing a transfected
`receptor (Amiranoff et al. 1984; Emami et al. 1986). GIP in-
`creases intracellular calcium in endothelial and some (Wheeler
`et al. 1995) but not all (Gremlich et al. 1995; Volz et al. 1995)
`cell lines transfected with the GIPR cDNA. A T340P mutation
`in the rat GIP receptor expressed in human embryonal kidney
`cells results in constitutive activation of adenylyl cyclase yet re-
`tention of GIP responsivity (Tseng and Lin 1997). Intriguingly,
`GIP also stimulates arachidonic acid production and release in
`Chinese hamster ovary cells expressing the rat GIPR, and in islet
`flTC-3 cells (Ehses et al. 2001).
`GLP-2 receptor activation is also coupled to activation of
`cAMP-dependent pathways in heterologous cell lines transfected
`with the rat or human GLP-2R (Munroe et al. 1999; Yusta
`et al. 1999). GLP-2 activates AP-1-dependent signaling path-
`ways, immediate early gene expression, and p70 S6 kinase ac-
`tivity, but does not stimulate intracellular calcium accumulation
`in BHK cells expressing the rat GLP-2R (Yusta et al. 1999). The
`GLP-2 effects on AP-1-dependent pathways are likely indirect,
`as GLP-2-stimulated induction of AP-1 luciferase activity was
`markedly attenuated in the presence of PKA inhibition (Yusta
`et al. 1999). GLP-2R signaling in BHK-GLP-2R cells is also
`coupled to activation of an antiapoptotic signaling program in
`a PKA-, phosphatidyl inositol 3-kinase-, and mitogen-activated
`protein kinase-independent manner (Boushey et al. 2001; Yusta,
`Boushey et al. 2000).
`
`RECEPTOR DESENSITIZATION
`The glucagon receptor undergoes both homologous and het-
`erologous desensitization in hepatocytes in vitro, and protein
`kinase C–selective inhibitors abrogate the heterologous desen-
`sitization process (Savage et al. 1995). Agonist occupancy of
`related hormone receptors, coupled to adenylyl cyclase in hep-
`atocytes, results in heterologous desensitization with kinetics
`
`Page 5
`
`
`
`184
`
`P. L. BRUBAKER AND D. J. DRUCKER
`
`than those observed for homologous
`slower
`somewhat
`desensitization (Premont and Iyengar 1988). Both homologous
`and heterologous glucagon receptor desensitization are indepen-
`dent of increases in intracellular calcium (Savage et al. 1995);
`receptor desensitization is also independent of cAMP activa-
`tion and is induced by activators of inositol phospholipid
`metabolism such as protein kinase C, independent of the in-
`hibitory guanine nucleotide regulatory protein Gi. (Murphy et al.
`1987, 1989).
`Prolonged activation of the GLP-1R by GLP-1 agonists
`results in a diminished response to subsequence ligand bind-
`ing in vitro (Fehmann and Habener 1991). Mutagenesis studies
`have identified three pairs of serine residues in the C-terminal
`tail of the receptor (S441/S442, S444/S445, and S451/S452)
`important for ligand-induced desensitization of the GLP-1R
`(Widmann et al. 1996a, 1996b, 1997). These amino acids are
`also necessary for internalization of the receptor, while phos-
`phorylation of an additional serine pair (S431/S432) by pro-
`tein kinase C appears to be involved in heterologous desensiti-
`zation of the GLP-1R (Widmann et al. 1996a, 1996b, 1997). It
`has been suggested that this heterologous desensitization may
`occur in the fl cell in response to ligand-mediated activation
`of phospholipase C, such as occurs in response to physiologi-
`cal regulators, including acetylcholine (Widmann et al. 1996b).
`Nevertheless, diminution of GLP-1 responsivity has not yet
`been clearly described in vivo, suggesting that other compen-
`satory mechanisms may serve to protect the receptor from
`downregulation.
`Homologous and/or heterologous desensitization of the GIPR
`may account for the reduced response to GIP that has been ob-
`served in patients with type 2 diabetes (Elahi et al. 1994; Nauck
`et al. 1993). The GIPR undergoes homologous desensitization
`(Fehmann and Habener 1991; Tseng, Boylan et al. 1996; Tseng
`and Zhang 2000), while incubation of islets with high glucose
`also reduces the response to GIP in a manner that is independent
`of protein kinases A and C (Hinke et al. 2000). The rat GIPR
`possesses only a single pair of serine residues in its C-terminal
`tail (S426/S427), and a combination of site directed mutagenesis
`and carboxyterminal deletions have implicated a role for these
`residues in receptor internalization, whereas S406 and C411
`contribute to receptor desensitization (Tseng and Zhang 1998;
`Wheeler et al. 1999). Cotransfection of a cDNA encoding fl-
`arrestin-1, or the G protein–coupled receptor kinase 2 (GRK2),
`but not GRK5 or GRK6, attenuated GIP-stimulated cAMP ac-
`cumulation in L293-GIPR and flTC3 cells (Tseng and Zhang
`2000). However, none of these proteins appear to be involved in
`GIPR internalization.
`In contrast to the information available for the glucagon,
`GLP-1, and GIP receptors, GLP-2 receptor desensitization has
`not yet been examined in cells expressing a GLP-2R in vitro.
`Although a diminished response to supraphysiological levels of
`GLP-2 is apparent in heterologous cells expressing the GLP-2R
`(Lovshin et al. 2001), mice treated with daily injections of GLP-
`2 for 12 weeks continued to exhibit enhanced intestinal mass,
`
`consistent with a lack of GLP-2R desensitization in vivo (Tsai,
`Hill, Asa et al. 1997).
`
`MOLECULAR BASIS OF LIGAND
`RECEPTOR-INTERACTION
`A series of N-terminal truncation and internal deletion mu-
`tants has been used to demonstrate that glycosylation is unlikely
`to be important for glucagon receptor intracellular trafficking
`and plasma membrane insertion (Unson et al. 1995). Remark-
`ably, deletion of the intracellular carboxyterminal tail of the rat
`receptor does not abrogate glucagon binding or adenylyl cy-
`clase generation in transiently transfected COS-1 or CHO cells
`(Bu