Minireview: Nutrient Sensing by G Protein-Coupled Receptors

Abstract

G protein-coupled receptors (GPCRs) are membrane proteins that recognize molecules in the extracellular milieu and transmit signals inside cells to regulate their behaviors. Ligands for many GPCRs are hormones or neurotransmitters that direct coordinated, stereotyped adaptive responses. Ligands for other GPCRs provide information to cells about the extracellular environment. Such information facilitates context-specific decision making that may be cell autonomous. Among ligands that are important for cellular decisions are amino acids, required for continued protein synthesis, as metabolic starting materials and energy sources. Amino acids are detected by a number of class C GPCRs. One cluster of amino acid-sensing class C GPCRs includes umami and sweet taste receptors, GPRC6A, and the calcium-sensing receptor. We have recently found that the umami taste receptor heterodimer T1R1/T1R3 is a sensor of amino acid availability that regulates the activity of the mammalian target of rapamycin. This review focuses on an array of findings on sensing amino acids and sweet molecules outside of neurons by this cluster of class C GPCRs and some of the physiologic processes regulated by them.

The common feature of the class C G protein-coupled receptors (GPCRs) discussed here, T1R1/T1R3, T1R2/T1R3, GPRC6A, and the calcium-sensing receptor (CaSR), is that each can detect amino acids. We begin by pointing out some unresolved questions concerning these receptors that are elaborated upon below. Natural ligands have been identified, but it seems possible that additional ligands may be found. These receptors have large Venus flytrap ligand-binding segments that can accommodate a variety of molecules, agonists, allosteric regulators, and inhibitors. A range of evidence indicates that these receptors are dimers: 2 are thought to be obligate heterodimers (T1R1/T1R3 and T1R2/T1R3) and 2 are thought to be homodimers (GPRC6A and CaSR). Although actions of these receptors are often consistent with signaling through the Gα_i family, the G protein-specific signal transduction mechanisms used by each receptor continue to be debated. The linkage of activation of these receptors to secretory responses and metabolic control is also in dispute. This review discusses roles of these receptors in pancreatic and gastrointestinal tissues. Although much is known, much also remains to be determined about the essential functions of these GPCRs.

T1R taste receptors

Identified as gustatory taste receptors and linked to a sweet-responsive locus in mice (1–3), the mammalian sweet and umami taste receptors were first characterized by Zuker and associates (4–6). T1R2 was demonstrated to be a heterodimer with T1R3 to form a receptor complex that is activated by a broad range of sweet compounds. The rodent and human T1Rs are only approximately 70% identical, which results in differences in agonist sensitivity (Table 1). For instance, in the initial studies, the human, but not rodent, T1R2/T1R3 was responsive to both aspartame and cyclamate. T1R2/T1R3 also responds to glycine and the sweet d-amino acids, but not their l-enantiomers. The T1R3 subunit was shown to heterodimerize not only with T1R2 but also with T1R1 to form the amino acid taste receptor complex (T1R1/T1R3). This heterodimer is responsive to most of the 20 l-amino acids, but not their d-enantiomers. Responses in a reconstituted system required expression of both subunits, consistent with the behavior of an obligate heterodimer. Studies looking at the purine nucleotide enhancement of the T1R1/T1R3 response to amino acids, monosodium glutamate, and the mGluR-agonist l(+)-2-amino-4-phosphonobutyric acid, led to the conclusion that T1R1/T1R3 was the umami receptor. Sequence differences between human and rodent T1R1 result in increased sensitivity of human T1R1 to glutamate (4). The connection of T1R1, T1R2, and T1R3 to both sweet and umami tastes was largely substantiated upon generation of mice in which the receptor genes were disrupted (6). Nerve activity, albeit decreased, was observed in response to sugars and umami compounds in T1R3 null mice developed in a different laboratory, suggesting that there may be additional taste mechanisms for these compounds (7). Perhaps such activity was mediated through related family members or alternative oligomeric receptor complexes (see Future Perspectives).

Table 1.

Nutrient and Pharmacological Regulators^a

Receptor	Agonists	Allosteric Activators	Inhibitors	References
T1R1/T1R3	l-Amino acids	Purine nucleotides	Lactisole	4, 100–103
	Monosodium glutamate	Cyclamate	Fibrates
		S809	Phenoxy herbicides
T1R2/T1R3	Fructose	Cyclamate	Lactisole	4, 100–104,114–116
	Galactose	S819	Fibrates
	Glucose	se2	Phenoxy herbicides
	Lactose	se3	Gurmarin
	Maltose
	Sucrose
	Gly
	d-Ala
	d-Asn
	d-Gln
	d-His
	d-Phe
	d-Trp Artificial sweeteners (e.g. Saccharin, Sucrolose, Aspartame)
GPRC6A	l-Amino acids	Ca2+	Calindol	56, 105, 106
	Osteocalcin		NPS2143
	Ca2+		2-Phenyl-indole-derived compounds 1–3
	Testosterone
CaSR	Ca2+	Amino acids	Calhex 231	31, 90, 107–110
		NPS R-568	NPS2143
		Calindol
		AMG073/Cinacalcet

^a There are selectivity differences between the human and rodent receptors. Please refer to References 4, 68, 101, and 110–114 for more information.

Many studies have suggested that stimulation of T1R1/T1R3 and T1R2/T1R3 in gustatory neurons leads to the activation of α-gustducin, a Gα_i family member associated with taste perception. Activation of α-gustducin and subsequent release of the Gβγ subunit are thought to cause an increase in phospholipase C (PLC) β-2 activity. Subsequent increases in inositol triphosphate induce intracellular Ca²⁺ release, leading to the gating of the transient receptor potential TRPM5 ion channel and subsequent depolarization (reviewed in Ref. 8). As early as 1996, the idea that α-gustducin might be involved in chemosenory mechanisms in nongustatory tissues was postulated after α-gustducin was found in stomach, intestine, and the pancreatic duct system (9, 10). A few years after discovery of the taste receptors, T1R2 and T1R3 were observed in mouse small intestine and human enteroendocrine l-cells and have since been evaluated as potential regulators of digestion and nutrient uptake (11). Dietary sugars increase intestinal glucose transport by increasing the expression of the gut sodium-dependent glucose co-transporter 1 (12–14). T1R3 and α-gustducin are necessary for increased sodium-dependent glucose co-transporter 1 induction by dietary carbohydrates and sucralose (15). Other studies have suggested that the activation of T1R1/T1R3 and T1R2/T1R3 by glutamate, glucose, and artificial sweeteners regulates the intestinal apical expression of a peptide transporter and the glucose transporter GLUT2 (16, 17). Thus, both heterodimers are thought to affect nutrient absorption.

Diverse studies have linked sweet taste receptors and α-gustducin to intestinal secretion of the amplifying hormone glucagon-like peptide 1 (GLP-1), which affects gut motility, increases satiety, and enhances insulin secretion from pancreatic β-cells in response to glucose. GLP-1 secretion induced by ingested glucose was reduced in mice null for α-gustducin; the deficit was also observed in isolated intestinal tissue. Secretion of GLP-1 induced by glucose, fructose, and sucralose from ileum explants from either T1R2 or T1R3 null mice was significantly reduced relative to that from wild type mice (18). Supported by T1R1 depletion experiments, activation of T1R1/T1R3 by phenylalanine, leucine, and glutamate in STC-1 enteroendocrine cells and primary mouse intestinal tissue induced the secretion of cholecystokinin (19). The T1R3 inhibitor lactisole inhibited glucose-induced GLP-1 secretion from human l-cells and was also reported to inhibit GLP-1 secretion in human subjects after intragastric or intraduodenal glucose infusion (20, 21). Ingestion of a diet soda sweetened with sucralose and acesulfame-K 10 minutes before an oral glucose tolerance test caused a significant increase in serum GLP-1 levels in human subjects in one study; however, no difference was seen in serum glucose or insulin concentrations between test and control groups (22). In a related study, administration of the artificial sweetener sucralose by direct infusion into the stomachs of human volunteers did not cause increased release of insulin, GLP-1, or gastric inhibitory peptide (23). A recent review provides an in-depth discussion on the effects of artificial sweeteners in the gastrointestinal tract (24). Part of the discrepancy in the studies with human subjects could be related to common polymorphisms in the T1R genes that affect the ability of these receptors to be activated by various ligands (25–28, 99). It is possible that these polymorphisms cause inconsistencies between subjects in their hormone secretion response to various artificial sweeteners. Therefore, future studies in individuals in which polymorphisms in T1R1, T1R2, and T1R3 genes are sequenced may be helpful in resolving these controversies. If these receptors enhance GLP-1 secretion, it remains to be determined whether it has a physiologic impact.

Agonist availability regulates the expression and/or membrane localization of taste receptors. Glucose or l-glutamate perfusion of rat jejunum for 20 minutes decreased the expression of T1R1, T1R2, and T1R3 at the apical membrane (16, 17). As blood glucose concentrations increased in type 2 diabetic human subjects, the expression of the receptors T1R2 and T1R3, and pathway components, TRPM5 and α-gustducin, decreased. In addition, the direct infusion of glucose into mouse jejunum resulted in a significant decrease in T1R2 expression (29). Expression of T1R1 and T1R3 was increased in the skeletal muscle of fasted mice compared with those of fed control mice (31). These findings support the idea that nutrients are relevant ligands for these receptors and suggest a feedback mechanism for receptor expression.

Although many studies have investigated the functions of T1Rs in the intestine, these GPCRs are expressed in many other tissues (31). All three T1R subunits and α-gustducin are present in mouse and human pancreatic islets and have been implicated in regulation of insulin secretion (31–33). Artificial sweeteners, sucralose, saccharin, and acesulfame-K, induce insulin secretion from rodent islets and a β-cell line (32). The activity of sweeteners was blocked by gurmarin, a peptide inhibitor of rodent T1R2/T1R3 (See Table 1) (15, 34). Fructose enhanced glucose-induced, but not basal, insulin secretion in mice and isolated mouse and human islets. This was not the case in T1R2^−/− mice, islets isolated from these mice, or human islets treated with the T1R3 inhibitor lactisole. A recent study reported that l-glutamate and l-arginine stimulate T1R1/T1R3 to promote insulin secretion; however, we find that this receptor affects insulin content more than insulin secretion, as discussed further below (31, 33).

Consistent with an action of taste receptors on glucose homeostasis, T1R3^−/− mice displayed delayed glucose clearance and decreased plasma insulin concentrations during an oral glucose tolerance test. However, T1R3^−/− and wild-type mice displayed similar plasma insulin concentrations following an ip glucose tolerance test, suggesting that it was T1R3 expression in the gut that was important for the modulation of plasma glucose. Insulin secreted from isolated islets from T1R3^−/− and wild-type mice was reported to be the same after 30 minutes of stimulation with 12.5 mM glucose; however, total internal reflection fluorescence microscopy revealed a lower rate of insulin granule fusion with secretory vesicles in islets from T1R3^−/−, but not T1R2^−/− mice (18). No differences in fusion kinetics were observed with 30 mM KCl as the stimulus. Still unclear is the exact role of these receptors in glucose homeostasis.

We investigated the role of T1R1/T1R3 in amino acid stimulation of the MAPKs ERK1/2 in pancreatic β-cells. ERK1/2 activation is a rapid indicator of the demand for insulin secretion in response to nutrients, hormones, and neurotransmitters. Once activated, ERK1/2 regulate nutrient-dependent insulin gene transcription (35). We found that amino acids activate ERK1/2 through T1R1/T1R3 because their activities were no longer increased by amino acids in cells depleted of one or the other receptor protomer. Amino acid-induced insulin secretion was reduced, but, as noted earlier in this paper, this was due to a significant loss of insulin content. Lowered insulin content might occur for a number of reasons. The mammalian target of rapamycin (mTOR) is controlled in most cells by amino acid availability. mTOR complex 1 (C1) inhibits autophagy and facilitates protein translation by phosphorylating and activating p70 S6 ribosomal protein kinase and phosphorylating and inhibiting eukaryotic initiation factor 4E-binding protein 1, a protein that binds and inhibits translation initiation by eukaryotic initiation factor 4E. Limiting amino acids reduces the activity of mTORC1, resulting in slowed translation, increased autophagy, and decreased cell growth. Consistent with a role for amino acid signaling through T1R1/T1R3 to mTORC1 in regulation of insulin content, depletion of T1R1 or T1R3 blocked mTORC1 activation by amino acids (Figure 1).

Figure 1.

Signaling through T1R1/T1R3, T1R2/T1R3, GPRC6A, and CaSR in β- or Gut Cells. In light green are GPCRs that can be directly activated by amino acids and use calcium as an allosteric regulator (A–C), whereas in blue is the CaSR whose main ligand is calcium and can be allosterically regulated by amino acids (D). See Table 1 for a comprehensive list of agonists, allosteric activators, and inhibitors for the reviewed GPCRs. A, T1R1/T1R3 receptor. In β, heart, and HeLa cells, the receptor activates PLCβ, ERK1/2, and mTORC1 and promotes release of calcium stores. Together, these pathways positively regulate 5′-cap-dependent transcription of the insulin message and may also promote insulin secretion. Receptor knockdown cells display reduced insulin levels and rampant autophagy. B, T1R2/T1R3 receptor. In enterocyte l-cells, the receptor activates PLCβ and promotes GLP-1 secretion and increased plasma membrane expression of the glucose transporters SGLT1 (sodium-dependent glucose co-transporter 1) and GLUT2 (glucose transporter 2). Signaling in β-cells is likely conserved and is proposed to regulate insulin secretion. C, GPRC6A receptor. In enterocyte l-cells the receptor activates PLCβ and promotes secretion of GLP-1. Signaling in β-cells is likely conserved and is proposed to regulate insulin secretion. D, CaSR receptor. In β-cells the receptor activates PLCβ and ERK1/2, induces release of ER calcium stores, stimulates expression of E-cadherin and l-type VDCC (black dotted arrows), inhibits Kir4.2 channels through protein-protein interactions, and promotes insulin secretion. The CaSR also regulates hormone secretion in other endocrine tissues. *, T1R2/T1R3 displays differential sensitivity for sweet compounds between humans and mice (4–6). †, Regulation of autophagy by the T1R1/T1R3 receptor was shown in heart and HeLa cells but not yet in β-cells. ‡, Described in enterocyte l-cells of the small intestine. See text and Table 1. ER, endoplasmic reticulum; IP₃, inositol triphosphate; IP₃R, IP₃ receptor.

Because mTORC1 is regulated by amino acids in most or all cell types, we examined expression of T1R1 and T1R3 in mouse tissues and several cultured cell lines and found that both were expressed in all sources tested. Thus, we asked whether T1R1/T1R3 is used for amino acid activation of mTORC1 more broadly throughout the body. Depletion of either protomer from cardiomyoblasts, HeLa, and MIN6 cells reduced mTORC1 activity, consistent with the observed decreased levels of mTORC1 activity in T1R3^−/− mouse tissues. In addition, receptor knockdown cells were deficient in recruitment of translation initiation complexes to 5′-capped messages (31). Autophagy, a process of degradative turnover of proteins and organelles that provides cells with nutrients to survive during conditions of nutrient limitation, is usually inhibited by mTORC1. We found that autophagy was increased as T1R1 or T1R3 expression was decreased even in the presence of amino acids. Counter to the dogma in the field which posited that mTORC1 only detects amino acids after they enter cells, we conclude that amino acid availability is first reported to mTORC1 by a T1R1/T1R3-mediated process.

Taste receptors generally signal through the Gα_i protein α-gustducin, yet α-gustducin knockout mice are still able to sense sweet and umami flavors (36–39). In addition, in some tissues T1R1/T1R3 has been proposed to signal through Gα_s or the Gα_q family member Gα₁₄ (32, 40, 41). In vitro reconstitution of G protein-coupled signaling using purified G proteins and purified 7-transmembrane-spanning core domains of T1R1, T1R2, and T1R3 showed that T1R1 and T1R2 activate Gα_i/o family members but not Gα_s or Gα_q, whereas T1R3 was unable to activate any of the tested Gα-subunits (42). These in vitro data support a preference of the 7-transmembrane-spanning core domains for Gα_i-subunits but fails to address how full-length heterodimeric sweet and umami receptors, as well as homodimeric T1R3 calcium-tasting receptors (43), signal in vivo to heterotrimeric G proteins.

In β-cells, Gα_i family proteins inhibit insulin secretion, whereas receptors that couple to Gα_s and Gα_q family members generally augment glucose-stimulated insulin secretion (44–47). We find that amino acid stimulation of starved β-cells activates ERK1/2 in a cAMP-independent and pertussis toxin-insensitive manner (Wauson, E.M., M. L. Guerra, K. McGlynn, E. Ross, and M. H. Cobb, in preparation), suggesting that in β-cells T1R1/T1R3 does not signal through Gα_i or Gα_s. Moreover, amino acid-induced activation of ERK1/2 and mTORC1 in β-cells was reduced by PLC inhibitors, calcium chelators, and calcium channel blockers (31). Together these findings indirectly suggest that mediation of amino acid activation of mTORC1 by T1R1/T1R3 may be through Gα_q but not Gα_i, despite the strong connection of taste receptors to α-gustducin in other systems. Because amino acids do not inhibit insulin secretion, Gα_i family members may not be coupled to amino acid receptors or inhibitory effects of Gα_i may be squelched by other signaling events. Signaling through Gβγ appears to be an important and dominant component in gustatory neurons and, like Gα_q but unlike Gα_i, also activates PLCβ (42, 48–50). The potential synergism between Gα_q and Gβγ in response to amino acid stimulation in β-cells should be explored.

Adding complexity to this system, amino acid-stimulated mTORC1 activity in C2C12 myoblasts was reduced independently both by treatment with pertussis toxin (Gα_i) and by knock down of a PLCβ isoform (Gβγ and/or Gα_q) (51). Together, these results indicate that the model of T1R1/T1R3 signaling, as defined in gustatory neurons, is unlikely to be generalized to all tissues. There is a gap in our understanding of the dominance and synergistic/antagonistic coupling of signals between Gα subunits and between Gα and Gβγ subunits in nongustatory tissues. With regard to β-cells, it still needs to be resolved whether signaling mechanisms activated by T1R1/T1R3 leading to different outputs are initiated by α-gustducin, Gβγ, Gα_q family members, or distinct G protein-dependent or -independent events, singly or in combination.

In surveying the diverging findings on responsiveness and functionality of the taste receptors, we suggest that differences in ligand specificity, organism- and cell type-selective coupling to signal transduction pathways, and receptor-dimer partnerships may complicate interpretation of their activities (see “Future Perspectives”).

GPRC6A

GPRC6A, like the T1R1/T1R3 heterodimer, can be activated by amino acids, and some of its actions on glucose homeostasis are reminiscent of taste receptor effects. This family C-GPCR was cloned independently by 2 separate groups (52, 53). Glycine and the l-amino acids, arginine, lysine, alanine, serine, ornithine, and citrulline, activated a chimeric GPRC6A receptor expressed in Xenopus oocytes (52, 54) (Table 1). The basic amino acids are the most potent agonists of wild-type mouse GPRC6A transfected into a human embryonic kidney (HEK)293 cell line (55). In addition to amino acids, GPRC6A is activated by calcium, other cations, calcimimetics, and osteocalcin (Ocn), a protein secreted by osteoclasts (56). GPRC6A is expressed in a pancreatic β-cell line and in mouse pancreas. By activating GPRC6A, Ocn may induce insulin secretion and insulin expression in islets (56) and GLP-1 secretion in intestinal L cells (57). The effect of Ocn on serum insulin was enhanced by the dipeptidyl peptidase IV inhibitor sitagliptin and decreased by exendin-(9–39), a GLP-1 receptor agonist, suggesting that the effect of Ocn on insulin concentration was through increased GLP-1 (57). This is consistent with GLP-1 secretion induced by ornithine from an intestinal L cell line (58). Cations including calcium and magnesium either activate GPRC6A directly, or positively modulate the receptor in the presence of amino acids (55, 59). GPRC6A may also be activated by testosterone and involved in nongenomic effects of steroid hormones (60). These cell-based studies suggest a role for this receptor in glucose homeostasis but the mechanism has not yet been clarified by characterization of GPRC6A^−/− mice.

GPRC6A^−/− mice were independently developed in 2 laboratories that have drawn somewhat different conclusions (61, 62) (Figure 1). In one study GPRC6A^−/− mice displayed the hallmarks of metabolic syndrome including insulin resistance, glucose intolerance, hyperglycemia, and hepatic steatosis, and also manifested a decrease in bone mineral density (61). These GPRC6A^−/− mice secreted significantly less insulin in response to l-arginine than wild-type mice, contained less pancreatic insulin, and had reduced pancreatic mass (63). The other group reported that the GPRC6A^−/− mouse neither exhibited a detectable reduction in bone mineral density, nor was there a defect in l-arginine-induced insulin secretion, either in vivo or from isolated pancreatic islets (62, 64). Additionally, no metabolic defects were observed in these mice on a normal diet. However, GPRC6A^−/− mice on a high-fat diet had significantly increased body weight, and increased circulating insulin and leptin concentrations compared with wild-type mice on the same diet (65). The more obvious metabolic abnormalities appeared in mice in which exon 2 of GPRC6A had been targeted, whereas exon 6 was targeted in the other study. These differences and potential animal background differences may have contributed to the penetrance of the phenotype. GPRC6A appears to be involved in glucose metabolism; whether its primary actions are on GLP-1 secretion or directly on insulin secretion appears to be an open question.

Extracellular Ca²⁺ sensing receptor

The extracellular CaSR is activated by calcium and other divalent and trivalent organic and inorganic cations (66). In gastric mucosa the CaSR also senses pH and l-amino acids, aromatic > aliphatic > polar amino acids (66, 67) (Table 1). The CaSR regulates a wide variety of cellular processes, including proliferation, differentiation, gene expression, ion channel function, and hormone secretion. The most studied roles are the homeostatic maintenance of systemic calcium through regulating synthesis and secretion of PTH, and the concentration of calcium in urine by the kidney (66).

The CaSR is widely expressed, including in pancreatic acinar and ductal cells, where it is reported to regulate secretion of digestive enzymes (68–70). In islets, it is expressed predominantly in α- and β-cells (71–75). CaSR is proposed to signal through Gα_i/o, Gα_q/11, Gα_12/13, and even Gα_s (66, 68, 76, 77). Knockdown and other studies suggest a requirement for PLC/inositol triphosphate and G_q/11 in β-cells (70, 72, 78, 79). ERK1/2 activation is also important for stimulation of proliferation by CaSR (78, 80, 81). Studies in β-cells are needed to determine which heterotrimeric G proteins couple to CaSR, and how it signals to ERK1/2.

Physiologic functions of CaSR in tissues that do not control systemic calcium are poorly understood. CaSR may monitor extracellular calcium in the exocrine pancreas to prevent formation of calcium stones and pancreatitis (69, 70). It has been proposed that gastrin-secreting cells of the gastric antrum use the CaSR to sense dietary calcium; activation of CaSR by amino acids was reported to induce the secretion of cholecystokinin and GLP-1 from intestinal cells (82–84). A general model discussed later in this paper suggests that the CaSR senses Ca²⁺ and other divalent cations in the extracellular space to synchronize and coordinate local calcium sensing and exocytosis (85) (Figure 1). This last hypothesis seems to fit findings in islet cells. CaSR stimulates glucagon secretion from islets (78, 86). In islets and β-cell lines, CaSR activation in the presence of substimulatory glucose promotes the rapid release of calcium stores into the cytosol and a robust first-phase insulin secretion. However, prolonged incubation with high extracellular calcium inhibits second-phase secretion (30, 72, 73, 75, 78, 81).

In considering the model suggested above, secretory granules are estimated to contain 60–120 mM calcium (87); recent measurements of calcium release into pseudoislet intercellular spaces ([Ca²⁺]_inter) indicate that within 15 minutes after stimulation with glucose, [Ca²⁺]_inter increases by about 1.31 mM (86). Such a concentration of calcium may be high enough to activate CaSR signaling. In support of this model, in fibroblasts cocultured with HEK293 cells, the local change in extracellular calcium following fibroblast exocytosis is sufficient to stimulate the CaSR on neighboring HEK293 cells (88). A recent study in MIN6 cells reports that calcium promotes expression of the epithelial adhesion protein E-cadherin and stimulates cell-cell tethering (80). CaSR expression appears to be regulated by its 3-dimensional distribution in the tissue; although MIN6 pseudoislets express one-third of the CaSR as MIN6 monolayers, pseudoislets respond more robustly to glucose stimulation compared with monolayers, and this response is decreased in CaSR-depleted pseudoislets (80, 81, 89). These findings suggest that islets and pseudoislets should be more sensitive to calcium than monolayers, that there should be CaSR-dependent crosstalk between α- and β-cells, and that CaSR signaling should be important for pancreatic islet formation.

The CaSR also modulates β-cell granule secretion by proximity to 2 different types of cation channels: l-type voltage-dependent calcium channel (VDCC) and potassium inward rectifying (Kir) channels. Calcium stimulation was reported to promote expression of the l-type VDCC (80), and 2 recent studies showed that both glucose and l-histidine promote colocalization of the CaSR with the l-type VDCC (90, 91). Although the physical details are lacking, this putative interaction is highly relevant because the l-type VDCC is an important component of insulin secretion. With regard to Kir channels, in kidney cells, stimulation of CaSR promotes its association with Kir4.1 and Kir4.2, and this association inactivates both potassium channels (92). A recent study in INS1 cells shows that the Kir4.2 channel is a negative regulator of glucose-stimulated insulin secretion, and that it interacts with the CaSR (93). Similar to the case in kidney cells, association with the CaSR may inhibit Kir4.2 function, which may be required for efficient glucose-stimulated insulin secretion. The potential regulation of l-type VDCC and Kir4.1/Kir4.2 channels by CaSR signaling in β-cells should be further explored. As with the GPCRs discussed earlier in this paper, CaSR appears to be intimately involved with glucose homeostasis.

Future Perspectives

The heterotrimeric G protein signaling downstream of the umami, sweet, CaSR, and GPRC6A receptors still needs to be defined in greater detail. It seems clear that extrapolation from characterized systems (eg, gustatory neurons) will be inadequate to explain signaling in all systems. It is likely that complex relations exist between Gα subunits as well as between Gα and Gβγ subunits that need to be explored.

Although much is known about CaSR desensitization and membrane trafficking between the plasma membrane and endolysosomal compartments (reviewed in Refs. 94 and 95), little is known about how umami, sweet, and GPRC6A receptors are desensitized, or how they are internalized and recycled/degraded in response to agonist stimulation.

These receptors are structurally similar and have a large common set of ligands and allosteric regulators (eg, amino acids and calcium; see Table 1). A recent study showed that T1R3, a protomer thought to be in obligate heterodimers, can homodimerize and act as an unexpected calcium taste receptor (43). This suggests that homo-, hetero-, and higher order oligomeric interactions among different family members should be explored (96, 97); such interactions introduce potential for unanticipated ligand regulation and novel outputs. Crosstalk between other members of the class C family of GPCRs has been reported, some of which involves formation of higher order oligomers (reviewed in Ref. 98).

Finally, it is self-evident that these class C receptors are promiscuous in their ligand selection (or, at the least, we have not recognized essential specificity determinants); rather they seem to respond to common ligands with a different order of selectivity (Table 1). Screens to determine what other nutrient arrays activate or modulate the activity of these combinatorial receptor states should provide a better understanding of how cells integrate nutritional cues to make appropriate decisions.