G-protein-coupled receptors (GPCR) represent the largest family of cell-surface mediators of signal transduction( Reference Milligan and McGrath 1 ). GPCR have attracted significant attention in terms of continued identification and characterisation, with recognition that they are targets for novel drug discovery. With more recent evidence demonstrating that nutrient sensing in the gastrointestinal tract is accomplished by a number of GPCR( Reference Wellendorph, Johansen and Bräuner-Osborne 2 ), the role of these receptors as important nutritional targets is becoming evident.
Nutrient-sensing GPCR for a variety of nutrients have been identified in the intestinal epithelium. They are expressed on the apical domain of enteroendocrine (sensor) cells of the gut and are directly activated by nutrients( Reference Dyer, Daly and Salmon 3 – Reference Hansen, Rosenkilde and Knop 9 ). Nutrient sensing initiates a cascade of events involving hormonal and neural pathways. This culminates in functional responses that ultimately regulate vital processes such as nutrient digestion and absorption, food intake, insulin secretion and metabolism.
This brief article focuses on the role of the taste receptor 1 family of GPCR, T1R1, T1R2, and T1R3, in sweet and l-amino acid sensing, with particular focus on its role in glucose absorption, glucose homeostasis and satiety. Moreover, the impact of exploiting the T1R2–T1R3 heterodimer as a nutritional target for enhancing intestinal glucose (salt and water) absorption and gut structural maturity in young animals is highlighted.
Sweet and l-amino acid sensing in the lingual epithelium
The T1R family present in the taste cells of the lingual epithelium consists of three members: T1R1; T1R2; T1R3( Reference Chandrashekar, Hoon and Ryba 10 , Reference Li, Staszewski and Xu 11 ). These receptors are distantly related to metabotropic glutamate receptors (mGluR), extracellular Ca2+-sensing receptor (CaSR) and γ-aminobutyric acid type B receptor( Reference Chandrashekar, Hoon and Ryba 10 ). Based on electrophysiological studies, heterologous expression of taste receptor subunits and behavioural assays of knockout mice, the heterodimeric combination of T1R2–T1R3 has been shown to function as a broad-specificity sweet sensor for natural sugars, sweet proteins and artificial sweeteners, whereas the combination of T1R1–T1R3 has been identified as a broad-spectrum l-amino acid sensor, responsible for mediating the perception of the savoury ‘umami’ taste of monosodium glutamate( Reference Li, Staszewski and Xu 11 , Reference Nelson, Chandrashekar and Hoon 12 ). Both the T1R2–T1R3 and T1R1–T1R3 heterodimers are coupled to the heterotrimeric G-protein gustducin to transmit intracellular signals( Reference McLaughlin, McKinnon and Margolskee 13 ).
In rodents and many other mammalian species, the lingual epithelium T1R1–T1R3 heterodimer responds to most of the twenty standard l-amino acids in the millimolar range( Reference Nelson, Chandrashekar and Hoon 12 ). However, the T1R1–T1R3 heterodimer is not activated by l-tryptophan (TRP)( Reference Nelson, Chandrashekar and Hoon 12 ). The human T1R1–T1R3 complex functions as a much more specific receptor, responding selectively to monosodium glutamate (GLUT) and aspartate (as well as to the GLUT analogue L-AP4)( Reference Chandrashekar, Hoon and Ryba 10 , Reference Li, Staszewski and Xu 11 ). A salient feature of amino acid taste in animals and umami taste in humans is the synergistic enhancement of potency when GLUT or other amino acids combine with the monophosphate esters of inosine or guanosine nucleotides (IMP and GMP)( Reference Yamaguchi 14 – Reference Yoshii, Yokouchi and Kurihara 16 ). Both GLUT and IMP/GMP bind to adjacent domains on the N-terminal Venus flytrap module of T1R1( Reference Zhang, Klebansky and Fine 17 ), while potentiation of intracellular signal transmission by IMP is mediated through α-gustducin( Reference He, Yasumatsu and Varadarajan 18 ). Gurmarin, a thirty-five-residue polypeptide from the Indian-originated tree Gymnema sylvestre (Gurmar), can inhibit both sweet and l-amino acid sensing by binding to the Venus flytrap domain of T1R3, inhibiting its function( Reference Imoto, Miyasaka and Ishima 19 – Reference Daly, Al-Rammahi and Moran 24 ).
Intestinal sweet sensing
Work carried out in many laboratories has demonstrated that T1R family members and gustducin are co-expressed in enteroendocrine cells in a range of species( Reference Jang, Kokrashvili and Theodorakis 5 , Reference Margolskee, Dyer and Kokrashvili 7 , Reference Daly, Al-Rammahi and Moran 24 – Reference Daly, Al-Rammahi and Arora 33 ), suggesting that taste-sensing mechanisms exist in the gastrointestinal tract.
It is well established that enteroendocrine L and K cells secrete glucagon-like peptides (GLP) (1 and 2) and glucose-dependent insulinotropic peptide (GIP), respectively, on encountering glucose in the intestinal lumen. GLP-1 and GIP, known as incretins, enhance insulin secretion, while GLP-2 increases intestinal growth and glucose absorption( Reference Rehfeld 34 – Reference Sangild, Tappenden and Malo 36 ). The infusion of intestinal lumen with the d-isoforms of glucose, galactose and fructose and non-metabolisable analogues of glucose, 3-O-methyl-glucose and α-methyl-glucose, causes the secretion of GIP and GLP-1 in rats, pigs and humans( Reference Dumoulin, Moro and Barcelo 37 – Reference Layer, Holst and Grandt 39 ). Furthermore, it has been shown that the T1R2–T1R3 heterodimer together with the α-subunit of gustducin resides in K and L endocrine cells containing GIP, GLP-1 and GLP-2, respectively( Reference Margolskee, Dyer and Kokrashvili 7 , Reference Moran, Al-Rammahi and Arora 8 , Reference Batchelor, Al-Rammahi and Moran 30 , Reference Daly, Al-Rammahi and Arora 33 ).
Functional evidence for the role of the T1R2–T1R3 heterodimer in intestinal glucose (sweet) sensing, inducing GLP-1, GLP-2 and GIP release, has been provided using endocrine cell lines, native intestinal tissue explants and knockout mice deficient in α-gustducin or T1R3( Reference Margolskee, Dyer and Kokrashvili 7 , Reference Daly, Al-Rammahi and Moran 24 , Reference Geraedts, Takahashi and Vigues 32 , Reference Daly, Al-Rammahi and Arora 33 ). The murine endocrine cell line GLUTag exhibits markedly increased GLP-1 secretion upon exposure to the artificial sweetener sucralose; this secretion is blocked by gurmarin, indicating that sucralose-induced GLP-1 release occurs through the activation of the T1R2–T1R3 heterodimer( Reference Margolskee, Dyer and Kokrashvili 7 ). Similar results were obtained for sucralose-induced GLP-1 release in the human L endocrine cell line NCI-H716, which was blocked either by RNA interference targeting of α-gustducin or by the human sweet taste receptor antagonist lactisole( Reference Jang, Kokrashvili and Theodorakis 5 ). Furthermore, the plasma levels of GLP-1 and GIP following the introduction of glucose directly into the proximal intestine are reduced in α-gustducin or T1R3 knockout mice, compared with wild-type controls( Reference Kokrashvili, Mosinger and Margolskee 40 ). Moreover, these knockout mice have abnormal insulin profile and prolonged postprandial blood glucose responses in response to luminal glucose( Reference Kokrashvili, Mosinger and Margolskee 40 ). Further work carried out by Geraedts et al. ( Reference Geraedts, Takahashi and Vigues 32 ) has shown that luminal glucose, fructose or sucralose evoke release of GLP-1 from mouse ileal explants embedded in an Ussing chamber, and that secretion of GLP-1 does not occur in tissue explants from T1R3 knockout mice( Reference Geraedts, Takahashi and Vigues 32 ). Moreover, exposure off mouse intestinal explants to either glucose or sucralose results in the secretion of GLP-1 and GLP-2, in a dose-dependent manner, and that this secretion is inhibited in the presence of gurmarin, a specific inhibitor of T1R3( Reference Daly, Al-Rammahi and Arora 33 ) (Fig. 1). Notably, the levels of GLP-1 and GLP-2 released by control and glucose-stimulated tissues were similar to those observed in in vivo studies in rats and human subjects given glucose orally or maintained as controls( Reference Brubaker, Crivici and Izzo 41 , Reference Smushkin, Sathananthan and Man 42 ), supporting the suitability of intestinal tissue explants for such studies. In these assays, the endocrine cells reside in their native niche, and it appears that maintaining contacts with neighbouring cells is important for endocrine cells to retain their functional viability( Reference Rogers, Tolhurst and Ramzan 43 ). Collectively, the data suggest that the sensing of sugars by the T1R2–T1R3 heterodimer coupled to gustducin expressed in L and K endocrine cells leads to the release of GLP-1, GLP-2 and GIP.
However, there are reports indicating that sweeteners do not trigger the release of incretins. Parker et al. ( Reference Parker, Habib and Rogers 44 ) have reported that primary cultures of adult mouse intestine do not secrete GIP in response to sucralose. This is not surprising, since they have indicated that these isolated cells do not express the T1R2–T1R3 heterodimer( Reference Parker, Habib and Rogers 44 ). There are also reports that oral ingestion or intragastric infusion of artificial sweeteners does not increase the secretion of incretins in rats( Reference Fujita, Wideman and Speck 45 ) or humans( Reference Ma, Bellon and Wishart 46 ). By feeding rats a single concentration of sweeteners (50 mg or 1 g/kg body weight, depending on the sweetener), Fujita et al. ( Reference Fujita, Wideman and Speck 45 ) have concluded that sweeteners do not acutely induce the release of incretin hormones. Ma et al. ( Reference Ma, Bellon and Wishart 46 ) have also reported that 0·4 or 4 mm-sucralose given by intragastric infusion does not induce the secretion of incretins. Interestingly, lactisole, which inhibits T1R3 function, reduces the blood levels of GLP-1 in humans receiving an intragastric glucose load( Reference Steinert, Gerspach and Gutmann 47 ).
Many artificial sweeteners are partly absorbed in the stomach and subsequently secreted in the urine( Reference Renwick 48 ). Therefore, the lack of response observed by these workers may be due to the concentration of the sweeteners being below the threshold level required for activating the candidate receptor and/or the lack of availability of the sweetener at the distinct target region of the intestine. Further work is required to unravel these controversies.
The majority of membrane-bound proteins, including GPCR, are low-abundance proteins( Reference Akermoun, Koglin and Zvalova-Iooss 49 – Reference Choksawangkarn, Kim and Cannon 53 ). In our experience more sensitive SYBR green assay rather than TaqMan-based assay and/or increased amounts of template complementary DNA (up to 250 ng/reaction) are effective in detecting the expression of T1R family members, having low abundance mRNA. This is perhaps why one or two laboratories have failed to detect the expression of T1R1, T1R2, T1R3 and gustducin in purified primary enteroendocrine cells using quantitative PCR( Reference Parker, Habib and Rogers 44 , Reference Reimann, Habib and Tolhurst 54 , Reference Oya, Kitaguchi and Pais 55 ). Other factors, such as the prevailing cell isolation conditions or the small proportion of purified L or K cells expressing taste receptor subunits and gustducin, have also been proposed to be responsible for the lack of detection of taste receptor elements in purified L and K cells( Reference Young 56 ).
There are some reports proposing that T1R subunits are expressed in the colon; however, their precise functions require further investigations. Iwatsuki et al. ( Reference Iwatsuki, Nomura and Shibata 57 ) have demonstrated the expression of T1R2–LacZ in mouse small and large intestinal tissues. Geraedts et al. ( Reference Geraedts, Takahashi and Vigues 32 ) have reported glucose-stimulated GLP-1 secretion from Ussing chamber-embedded large intestinal explants of T1R3, but not T1R2, knockout mice. They have concluded that T1R3-dependent and independent pathways are involved in the regulation of GLP-1 secretion in the colon( Reference Geraedts, Takahashi and Vigues 32 ).
It should be borne in mind that L cells in the small and large intestines may have different phenotypes. Furthermore, in the lumen of the native colonic tissue, there is hardly any free glucose available. Glucose is rapidly metabolised to SCFA by colonic microbiota. SCFA induce the release of GLP-1 via colonic endocrine L-cell GPR43 (FFAR2)( Reference Tolhurst, Heffron and Lam 58 ). Therefore, studies directed at the sensing of nutrients in the hindgut must always consider the digestive activity of the microbiota.
Mechanisms underlying intestinal sweet sensing and glucose transport regulation: application to the maintenance of gut health in weaning piglets
The major route for the absorption of dietary glucose (and galactose) from the lumen of the intestine into enterocytes is via the apical membrane Na+/glucose cotransporter-1 (SGLT1)( Reference Dyer, Al-Rammahi and Waterfall 59 – Reference Gruzdkov, Gromova and Grefner 62 ). The absorption of glucose by SGLT1 also activates salt (NaCl) and water absorption; this is used as the route for oral rehydration therapy( Reference Hirschhorn and Greenough 63 ). Thus, the regulation of SGLT1 is essential for the provision of glucose to the body and avoidance of intestinal malabsorption. A number of studies( Reference Moran, Al-Rammahi and Arora 60 , Reference Gorboulev, Schürmann and Vallon 61 , Reference Ferraris and Diamond 64 – Reference Stearns, Balakrishnan and Rhoads 67 ) have established that the expression of intestinal SGLT1 is enhanced in response to a range of monosaccharides, including non-metabolisable analogues of glucose. Furthermore, it has been shown that the pathway underlying monosaccharide-enhanced SGLT1 expression involves a luminal membrane GPCR glucose sensor( Reference Dyer, Vayro and King 66 ).
Convincing evidence for the involvement of gut-expressed T1R2–T1R3 heterodimer and gustducin in intestinal sweet transduction and SGLT1 regulation has been provided by studies using mice in which the genes for either α-gustducin or the sweet receptor subunit, T1R3, had been deleted. The elimination of sweet transduction in mice in vivo has been shown to prevent the dietary monosaccharide-induced up-regulation of SGLT1 expression that is observed in wild-type mice( Reference Margolskee, Dyer and Kokrashvili 7 ). Furthermore, it has been demonstrated that artificial sweeteners when included in the diet also enhance the expression of SGLT1( Reference Margolskee, Dyer and Kokrashvili 7 ). In cats (Felidae family) and chickens, naturally occurring ‘T1R2 knockout’ models, there is a good correlation between the absence of T1R2 expression and the inability to increase SGLT1 expression in response to increased dietary sugars( Reference Batchelor, Al-Rammahi and Moran 30 , Reference Buddington, Chen and Diamond 68 , Reference Barfull, Garriga and Mitjans 69 ). All together, the data support the notion that the T1R2–T1R3 heterodimer, in association with gustducin, senses dietary sugars to regulate the expression of intestinal SGLT1( Reference Margolskee, Dyer and Kokrashvili 7 ).
To unravel the underlying mechanism by which sugar activation by the T1R2–T1R3 heterodimer, expressed on the apical domain of endocrine cells, leads to the up-regulation of SGLT1 expression in neighbouring enterocytes, the underlying chemosensing mechanism has been investigated. It is well established that systemic infusion of GLP-2 enhances intestinal growth and SGLT1 expression( Reference Cottrell, Stoll and Buddington 35 , Reference Sangild, Tappenden and Malo 36 , Reference Tsai, Hill and Asa 70 – Reference Cheeseman 72 ). Moreover, it has been demonstrated that in vivo vascular infusion of GLP-2 increases, with a similar magnitude, the maximal rate of Na+-dependent glucose transport, Na+-dependent phlorizin binding and SGLT1 protein abundance in the intestinal brush border membrane. This GLP-2 effect was inhibited by brefeldin A( Reference Cheeseman 72 ), an inhibitor of protein translocation from the trans-Golgi apparatus to the plasma membrane( Reference Hunziker, Whitney and Mellman 73 – Reference Helms and Rothman 75 ), suggesting that GLP-2, increases the number if SGLT1 protein molecules in the brush border membrane( Reference Cheeseman 72 ).
As shown in Fig. 1, the exposure of mouse small intestinal explants to glucose or sucralose evokes the secretion of GLP-2, in a dose-dependent manner, which is inhibited in the presence of gurmarin, indicating that glucose/sucralose-induced GLP-2 release occurs via the activation of the T1R2–T1R3 heterodimer. Since the GLP-2 receptor is expressed in enteric neurons( Reference Bjerknes and Cheng 76 ), and not in absorptive enterocytes, a direct paracrine effect of GLP-2 on the neighbouring enterocytes is excluded. The knowledge that direct administration of GLP-2 to enteric neurons induces a neuronal response( Reference Bjerknes and Cheng 76 , Reference Mills and Gordon 77 ) and that electric stimulation of enteric neurons results in the up-regulation of SGLT1 expression, which is inhibited by nerve blocking agents (our own observation), implies that the binding of GLP-2 to its receptor in enteric neurons stimulates a reflex response that results in increased functional expression of SGLT1 in absorptive enterocytes.
Impact
With an intensive livestock production, a shorter suckling period increases productivity in terms of numbers of piglets born. However, early weaning has adverse effects on the intestinal function of piglets, leading to nutrient malabsorption, diarrhoea, malnutrition and dehydration( Reference Everts, van Beers-Schreurs and Vellenga 78 – Reference Nabuurs, Hoogendoorn and van Zijderveld-van Bemmel 80 ). A number of field trials (involving more than 4500 piglets) have shown that artificial sweeteners, included in piglet feed, are effective in preventing post-weaning intestinal disorders, enhancing the growth and well-being of early-weaned piglets( Reference Sterk, Schlegel and Mul 81 ). It is notable that despite the increased palatability of feed containing artificial sweeteners, no steady increase in feed intake has been observed. However, a consistent enhancement of feed conversion efficiency (i.e. kg body mass gained per kg feed intake) has been observed, and the reason for this, until recently, was unknown. The understanding of the molecular basis by which artificial sweeteners enhance gut structural maturity and increase intestinal glucose (salt and water) absorption has led to an effective utilisation of sweeteners as dietary supplements, routinely included in the diet of early-weaned piglets to prevent post-weaning intestinal disorders.
Intestinal sensing of l-amino acids
Protein hydrolysates, peptides and amino acids elicit the secretion of cholecystokinin (CCK) both in vivo and in vitro ( Reference Konturek, Radecki and Thor 82 – Reference Sufian, Hira and Asano 91 ). CCK plays a variety of roles in digestive processes, such as slowing of gastric emptying, mediation of intestinal motility and stimulation of pancreatic and gall bladder secretions( Reference Dockray 92 – Reference White, Schwartz and Moran 95 ). It also inhibits food intake in a manner consistent with a role in satiety( Reference Moran 96 ). Amino acids, in particular l-phenylalanine (PHE), at physiological concentrations (10–50 mmol/l)( Reference Hira, Nakajima and Eto 97 , Reference Liddle, Walsh and Dockray 98 ) increase plasma CCK levels and reduce food intake in humans, monkeys, dogs and rodents( Reference Anika, Houpt and Houpt 99 – Reference Owyang, May and Louie 102 ). Leucine (LEU), a branched-chain amino acid, induces the release of CCK in cats( Reference Backus, Howard and Rogers 103 ).
T1R1 and T1R3 are expressed in mouse intestinal tissue( Reference Daly, Al-Rammahi and Moran 24 , Reference Dyer, Salmon and Zibrik 26 , Reference Wang, Inoue and Higashiyama 31 ) and in mouse enteroendocrine STC-1 cells( Reference Daly, Al-Rammahi and Moran 24 ). Immunohistochemistry, using triple immunolabelling, has demonstrated co-expression of T1R1, T1R3 and CCK in the same endocrine cell in the mouse proximal intestine( Reference Daly, Al-Rammahi and Moran 24 ). Furthermore, confocal microscopy has shown the expression of T1R1/T1R3 to be confined to the apical region, with CCK residing at the basal domain of the same endocrine cells. Immunohistochemical localisation, using double immunolabelling, of mouse proximal intestinal serial sections has confirmed that T1R1 is not expressed by S, K or L endocrine cells and that T1R1 expression is confined to CCK-containing I cells( Reference Daly, Al-Rammahi and Moran 24 ). The endocrine cells containing CCK also possess T1R1, T1R3 and α-gustducin( Reference Daly, Al-Rammahi and Moran 24 ).
Functional evidence for the role of the T1R1–T1R3 heterodimer in intestinal l-amino acid sensing and eliciting CCK release has been provided by using the STC-1 cell line and mouse proximal intestinal explants. The exposure of STC-1 cells to the individual l-amino acids PHE, TRP, LEU and GLUT provokes the secretion of CCK( Reference Daly, Al-Rammahi and Moran 24 ). In contrast, the d-isoforms of these amino acids have no effect, providing supportive evidence for the specific effect of l-isoforms on the induction of CCK secretion. Furthermore, the inhibition of T1R1 expression in STC-1 cells by RNA interference leads to a significant decrease in CCK secretion in response to PHE, LEU and GLUT, but not to TRP( Reference Daly, Al-Rammahi and Moran 24 ). TRP is a high potency activator of CaSR( Reference Conigrave, Quinn and Brown 104 ), but inactive for the T1R1–T1R3 heterodimer( Reference Nelson, Chandrashekar and Hoon 12 ). IMP, the specific potentiator of the T1R1–T1R3 heterodimer, significantly enhances the release of CCK by STC-1 cells in response to PHE, LEU and GLUT, but not to TRP. Moreover, pre-incubation of STC-1 cells with gurmarin inhibits the secretion of CCK significantly in response to PHE, LEU and GLUT, but has no effect on TRP-induced CCK release( Reference Daly, Al-Rammahi and Moran 24 ), collectively indicating that the T1R1–T1R3 heterodimer functions as a sensor for PHE-, LEU- and GLUT-induced CCK release in STC-1 cells.
Mouse proximal intestinal explants secrete CCK in response to PHE, LEU and GLUT and this secretion is enhanced by the addition of IMP. However, IMP has no effect on TRP-induced CCK secretion. Moreover, the release of CCK in response to PHE, LEU and GLUT, but not to TRP, is inhibited dramatically by pre-incubation of the tissue with gurmarin( Reference Daly, Al-Rammahi and Moran 24 ). Therefore, the functional properties and cellular location of gut-expressed T1R1–T1R3 heterodimer support its role as a luminal sensor for l-amino acid-induced CCK secretion in mouse proximal intestine.
Using isolated and purified mouse mucosal enhanced green fluorescent protein-expressing CCK cells, Wang et al. ( Reference Wang, Chandra and Samsa 105 ) and Liou et al. ( Reference Liou, Sei and Zhao 106 ) have shown that aromatic amino acids l-PHE and l-TRP stimulate the release of CCK through CaSR( Reference Wang, Chandra and Samsa 105 , Reference Liou, Sei and Zhao 106 ). We have shown that the addition of a CaSR antagonist, NPS2143, inhibits PHE-stimulated CCK release partially and TRP-induced CCK secretion totally in mouse proximal intestinal tissue explants, with no effect on LEU- or GLUT-induced CCK secretion (see Fig. 2). The partial and total inhibition of CaSR-mediated PHE- and TRP-induced CCK secretion is consistent with data presented by Wang et al. ( Reference Wang, Chandra and Samsa 105 ), using purified CCK–enhanced green fluorescent protein cells in the presence and absence of another CaSR antagonist, Calhex 231( Reference Wang, Chandra and Samsa 105 ).
Therefore, it appears that both receptors T1R1–T1R3 and CaSR are capable of sensing L-PHE. Interestingly, in support of this, the addition of NPS2143 together with gurmarin totally inhibits PHE-induced CCK release from mouse proximal intestinal tissue( Reference Daly, Al-Rammahi and Moran 24 ) (see Fig. 2). The experimental data suggest that CaSR acts as an intestinal amino acid receptor specifically sensing l-aromatic amino acids, while the T1R1–T1R3 heterodimer responds to a number of amino acids provoking CCK secretion.
Nutrient sensing GPCR are attractive and orally accessible targets for manipulations by functional foods and supplements. This has applications for maintaining health and preventing disease.
Acknowledgements
S. P. S.-B. acknowledges the University of Liverpool, the Wellcome Trust and Pancosma for providing financial support. The authors thank peer reviewers for providing constructive and critical comments. The authors' contributions are as follows: K. D., M. A.-R. and A. W. M. carried out the experiments; D. B. provided scientific and practical advice; S. P. S.-B. wrote the article. S. P. S.-B., K. D., M. A.-R. and A. W. M. declare no conflicts of interest. D. B. is an employee of Pancosma.