Taste receptors of the gut: emerging roles in health and disease
- Correspondence to Professor Inge Depoortere, Gut Peptide Research Lab, Translational Research Center for Gastrointestinal Disorders, University of Leuven, Gasthuisberg O&N1, box 701, Leuven 3000, Belgium;
- Received 29 July 2013
- Revised 16 September 2013
- Accepted 25 September 2013
- Published Online First 16 October 2013
Recent progress in unravelling the nutrient-sensing mechanisms in the taste buds of the tongue has triggered studies on the existence and role of chemosensory cells in the gut. Indeed, the gastrointestinal tract is the key interface between food and the human body and can sense basic tastes in much the same way as the tongue, through the use of similar G-protein-coupled taste receptors. These receptors ‘taste’ the luminal content and transmit signals that regulate nutrient transporter expression and nutrient uptake, and also the release of gut hormones and neurotransmitters involved in the regulation of energy and glucose homeostasis. Hence, they play a prominent role in the communication between the lumen, epithelium, smooth muscle cells, afferent nerve fibres and the brain to trigger adaptive responses that affect gastrointestinal function, food intake and glucose metabolism. This review summarises how sensing of nutrients by taste receptors along the gut plays a key role in the process of digestion, and how disturbances or adaptations of these chemosensory signalling pathways may contribute to the induction or resolution of a number of pathological conditions related to diabetes, obesity, or diet-induced symptom generation in irritable bowel syndrome. Targeting these receptors may represent a promising novel route for the treatment of a number of these diseases.
The gut tastes what we eat—bitter, sweet, umami, fat—in much the same way as the tongue through the use of similar taste receptors and chemosensory signalling pathways
In health, taste receptors sense nutrients from either a luminal or blood-borne direction and transmit signals that control the secretion of gut hormones and the expression of nutrient transporters to maintain energy and glucose homeostasis and gastrointestinal function
In disease, disturbances or adaptations in the expression or sensitivity of these taste receptors and their signalling pathways may affect digestive behaviour and metabolism
This is a new and emerging field, future studies aimed at a better understanding of how the sensing of nutrients in the gut is finely tuned by different taste receptors in health and disease, should help us to define novel drug targets
We choose to eat for many reasons, including to satisfy our hunger, to invite pleasant sensory experiences, and as a social ritual. However, we need to eat to acquire nutrients essential for life and health. Not surprisingly, the digestive system contains a diverse array of detectors that help to regulate ingestive decisions, impact nutrient assimilation, avoid or neutralise toxins, and elicit complex neural and endocrine responses that affect metabolism, gastrointestinal (GI) transit, satiation and satiety. Strikingly, many of the G protein-coupled receptors (GPCRs) that detect nutrients and toxins in the oral cavity and function as taste receptors there, also subserve important functions throughout the GI tract. Activation of these receptors triggers the release of neurotransmitters (eg, ATP) that will excite primary sensory afferent fibres and interact with neighbouring presynaptic cells to relay information to the hindbrain.1 A similar system of highly orchestrated interactions also operates in the gut and further points to the functional similarity of lingual and intestinal cells. I will review the state of our knowledge about taste receptor function along the entirety of the alimentary canal and discuss the implications of these functions for human physiology and disease.
Taste receptors and gustatory function
The taste system detects compounds that elicit at least five perceptual qualities: sweet, umami, bitter, sour and salty. The latter two are largely mediated by ion channels, and will not be discussed further here. By contrast, sweet, umami and bitter-tasting stimuli are detected by members of two GPCR families, the taste 1 receptor family (TAS1R) and the taste 2 receptor family (TAS2R), that are expressed in subpopulations of taste bud cells (figure 1). Subtypes of the TAS1R family heterodimerise to detect sweet (TAS1R2-TAS13) and umami (TAS1R1-TAS1R3).2 ,3 Animals that lack these subunits are deficient in their ability to taste umami and/or sweet stimuli.4 The presence of multiple binding sites on these receptors can explain the synergistic effects of 5′-ribonucleotides and glutamate in umami taste, or the ability of many chemically diverse compounds–including sugars, non-caloric sweeteners, D-amino acids and some proteins–to elicit a sweet taste. Consistent with their proposed evolutionary need to detect foods rich in nutrients, the umami and sweet receptors are low-affinity receptors, with EC50s in the millimolar range for their most common natural ligands.
Alternatively, we are quite sensitive to bitter-tasting compounds. While this is important for avoiding potential toxins, it also reduces the oral tolerance for many common pharmaceuticals.5 Bitter tastants are recognised by TAS2Rs.6 The human genome encodes 25 different TAS2Rs, which fall into three functional categories: specialists that respond to one or two compounds; generalists that are highly promiscuous in their stimulus responses; and those that are intermediate in their selectivity.7 Many compounds can activate more than one receptor, and activation thresholds can vary from millimolar to nanomolar levels. TAS2R polymorphisms are quite common in human populations, and can have pronounced effects on the ability of individuals to detect certain bitter compounds.8
Other GPCRs have been implicated in the detection of gustatory stimuli. For example, metabotropic glutamate receptors, specifically mGluR1 and mGluR4, may contribute to amino acid taste.9 The Ca2+-sensing receptor (CaSR) and GPRC6A have been proposed to mediate Ca2+ taste and also amino acid taste in extraoral tissues such as the gut.10 ,11 Two receptors for long-chain fatty acids (LCFAs), FFAR1 (a.k.a., GPR40) and GPR120 have been localised to taste cells and contribute to orosensory responses to fats (whether or not fat can be considered as the sixth taste is the subject of ongoing investigation).12 Additionally, receptors for short-chain fatty acids (SCFA), FFAR2 and FFAR3, have been detected outside the oral cavity.11 An as yet unidentified receptor may mediate the taste of polysaccharides. An overview of the different taste receptors involved in nutrient sensing is represented in table 1.
TAS1Rs and TAS2Rs are found in Type II taste cells, a morphological class of microvillar taste bud cells. Activation of a gustatory G-protein (eg, gustducin) coupled to these GPCRs results in the activation of phospholipase C β2 leading to inositoltriphosphate-mediated release of intracellular Ca2+ and activation of the transient receptor potential cation channel M5 (TRPM5) (figure 1). These events induce membrane depolarisation, action potential generation, and the release of ATP, which acts on purinergic receptors to activate gustatory afferents leading to activation of brain centres involved in taste perception. Sweet, umami and bitter taste receptors are differentially expressed within the Type II taste cells.13 This segregation is a key component of taste quality coding and hedonic discrimination in the gustatory periphery. The extent to which these various taste transduction components are colocalised in gut cells is unclear, but surely has consequences for how they impact GI function.
Role of taste receptors in nutrient detection and response during digestion
‘Gut’ peptides produced in taste cells of the tongue
Taste cells also express a number of bioactive peptides best known for their roles in metabolism, feeding and satiety, including: glucagon-like peptide-1 (GLP-1), glucagon, neuropeptide Y, peptide YY (PYY), cholecystokinin (CCK), vasoactive intestinal peptide and ghrelin.14 The functions of these peptides in taste buds are not fully understood, but at least some may act to modulate the responsiveness of the peripheral gustatory apparatus to certain taste stimuli. Interestingly, stimuli representing different taste qualities promote the release of different peptide combinations, suggesting that these peptides could also play a role in quality coding.15
Cognate receptors for all the peptides expressed in taste cells can be found on taste cells or on the adjacent afferent nerve fibres.14 Therefore, these peptides almost certainly have autocrine and paracrine functions within the taste bud and, in some cases, could function as cotransmitters with ATP. However, it is unclear whether taste bud peptides also reach the bloodstream and exert endocrine effects. Furthermore, peptide receptors expressed on taste cells may be targets of circulating peptides produced in the gut, adipose tissue, or other tissues. For example, systemic leptin has been shown to decrease sweet taste responsiveness through actions on leptin receptors present on a subset of taste cells.16 Thus, postingestive nutrient responses could feed back onto the peripheral gustatory apparatus to modulate its responsiveness to subsequent foods.
Nutrient sensing in the stomach
Upon ingestion of food, the proximal stomach relaxes to accommodate large amounts of food without increasing intragastric pressure. Tension-sensitive mechanoreceptors are activated by the arrival of food and relay their information via the vagus to the hindbrain to induce sensations of satiation.17 Nutrient sensing in the stomach has been considered to be a less important factor, but the role of chemosensory receptors should not be ignored. Indeed, recent studies showed that intragastric administration of the bitter agonist, denatonium benzoate (DB), inhibited gastric accommodation to nutrient drink infusion and tended to increase satiation scores.18 The contractile response induced by DB was mimicked in vitro in smooth muscle strips from the mouse fundus and was partially mediated via the gustatory G-protein subunit, α-gustducin, and involved the release of Ca2+ from intracellular and extracellular stores.19
The proximal stomach is also the major site of production of the orexigenic hormone, ghrelin, which is considered as a physiological meal initiator.20 How the reduction of plasma ghrelin during a meal might participate in the induction of satiation is incompletely explored. The fact that ghrelin levels are suppressed strongly by ingested proteins, weakly by proteins and biphasically by carbohydrates, suggests that the ghrelin P/D1 cells contain the machinery to sense nutrients.21 Recent immunofluorescence studies confirmed that the ghrelin cell is colocalised with the gustatory G-proteins, α-gustducin and α-transducin, the sweet and umami receptor subunit, TAS1R3, and the free fatty acid-sensing receptor, GPR120.22–24 It remains to be investigated whether the postprandial suppression of ghrelin by nutrients is indeed mediated via these receptors. However, the first functional studies showed that α-gustducin is involved in the effect of bitter compounds on ghrelin release and in sensing of medium-chain fatty acids in the diet necessary for the octanoylation of ghrelin.22 ,23 Additionally, GPR120 seems to play a role in the lipid-sensing cascade of the ghrelin cell.23 The distribution of taste receptors involved in nutrient sensing in endocrine cells along the gut is summarised in figure 2.
The arrival of food in the stomach stimulates the gastric phase of acid, pepsinogen and mucus production. The two principal triggers are distension of the stomach, which initiates local myenteric and vago-vagal reflexes, and the chemical content of the food. In the stomach, sensing of protein breakdown products is important as they activate the release of two important regulators of pepsinogen and acid secretion: gastrin (G-cells) and somatostatin (D-cells). Indeed, peptone has been shown to affect gastrin and somatostatin secretion in rats.25 Recent studies showed that mouse and pig antral G-cells and a subpopulation of D-cells express GPR92, a receptor activated by dietary protein hydrolysates.26 In both endocrine cell types, GPR92 is coupled to different signalling pathways. It is therefore conceivable that GPR92 may play an important role in adjusting the release of gastrin and somatostatin in response to protein digests in the chyme. In addition, G and D-cells express the amino acid taste receptor GPRC6A and the CaSR that act in concert with each other to sense a broad spectrum of amino acids ranging from basic and small neutral (GPRC6A) to aromatic (L-Phe) amino acids and Ca2+.27 ,28 The stimulation of gastrin after gavage with L-Phe, peptone, Ca2+ and a neutralising buffer was abolished in CaSR−/− mice suggesting a regulatory role for this receptor in gastrin release.29 The TAS1R family is not important for G-cell sensing of protein and amino acids since TAS1R3 is not expressed on G-cells and the gastrin response to peptone is not abolished in TAS1R1/3−/− mice.28 ,29 Additionally, L-Phe can induce acid secretion independent of hormonal stimulation via activation of the CaSR on parietal cells.30
In the distal stomach, the chyme is mixed with the digestive secretions and ground by powerful peristaltic contractions. The process of emptying starts and the rate is dependent on the gastric volume and the chemical nature of the gastric contents. The major control mechanism for gastric emptying involves duodenal gastric feedback and the emphasis shifts towards satiety mechanisms.
When the meal is emptied from the stomach, the upper GI motility pattern changes from a digestive to an interdigestive pattern: the migrating motor complex, characterised by a group of strong phasic contractions migrating distally. Phase 3 contractions, the most vigorous contractions of the migrating motor complex (MMC), originating in the stomach, are considered as a hunger signal. Arrival of nutrients will disrupt this pattern. Infusion of high doses of amino acids in healthy volunteers shortens the duration of the MMC length and suppresses antral phase 3 activity.31 Intragastric administration of the bitter compound, DB, in the fasted state was shown to decrease antral but not duodenal motility and to shift the origin of phase 3 from the antrum to the duodenum. This was accompanied by a significant decrease in hunger scores.32 In both cases, the mechanisms involved are unknown but may involve activation of taste receptors on endocrine cells, smooth muscle cells and/or nerves. Whether any stimulus that suppresses or induces gastric phase 3 activity inhibits or stimulates hunger, respectively, requires further investigation. Furthermore, recent studies showed that gastric phase 3 activity was less frequently observed in obese patients, and was associated with fewer hunger peaks suggesting that the absence of generation of gastric MMCs may be pro-anorexigenic and may also represent a compensatory mechanism to reduce hunger feelings in obesity.33 The motilin agonist, erythromycin, known to induce gastric phase 3, restored the association between gastric phase 3 and hunger peaks in obese patients.
Nutrient sensing in the small intestine
The small intestine is the major site of digestion and absorption in the GI tract. In the small intestine, carbohydrates, fats and protein digestion products, as well as osmolarity changes and physical distention activate inhibitory neural and endocrine pathways that signal the stomach to delay emptying and to help limit ingestion by enhancing gastric mechanoreceptor stimulation. However, experiments demonstrated that intestinal nutrient infusions inhibit food intake during sham feeding when gastric contents were drained via an open gastric cannula to obviate the gastric filling effects on food intake.34 Thus, while intestinal signals and gastric emptying interact to control food intake, a delay of gastric emptying is not required for intestinal signals to elicit satiation. The small intestine also sends endocrine satiety signals to the hypothalamus, either directly, via the bloodstream and across the incomplete blood-brain barrier to the arcuate nucleus, or indirectly, through the activation of the vagus nerve. It is clear that the vagal-brainstem-hypothalamic pathway plays a major role in the effect of the satiation hormones, CCK and GLP-1, on food intake. Their receptors are expressed on vagal afferents, they increase vagal afferent firing and induce cfos expression in the nucleus of the solitary tract after intraperitoneal injection.35–39 Furthermore the inhibition of food intake by exogenous GLP-1 and CCK is blocked by vagotomy and brainstem-hypothalamic pathway transectioning.40 ,41
In the duodenum, fat, protein hydrolysates and amino acids stimulate the secretion of CCK from I-cells. This hormone then acts to slow gastric emptying and increase satiety. In STC-1 cells, a mouse enteroendocrine-like cell line, LCFAs induce CCK secretion through the lipid sensor GPR120 and the transduction channel TRPM5.42 ,43 However, cell lines are incompletely validated as accurate models of native I-cells. Establishing cultures from endocrine cells remained a challenge for many years but the generation of transgenic mice with I cell-specific expression of a fluorescent protein, allowed the isolation of a pure I-cell population by fluorescence-activated cell sorting. Native I-cells are also immunoreactive for FFAR1, and the effect of linolenic acid on the secretion of CCK from isolated I-cells was abolished in cells from FFAR1−/− mice.44 ,45 Thus, GPR120 and FFAR1 appear to mediate lipid-induced CCK secretion.
Amino acids and protein hydrolysates similarly appear to use multiple receptor types to stimulate CCK secretion. In humans, administration of L-Phe increased CCK secretion and reduced food intake.46 Two amino acid-responsive receptors, TAS1R1-TAS1R3 and the CaSR, are expressed in I-cells, and implicated in CCK release.47 For example, Phe-dependent, Leu-dependent and Glu-dependent CCK release is dependent on TAS1R1-TAS1R3, while the CaSR mediates Phe-induced and Trp-induced responses in mouse intestinal tissue explants.47 Studies in vitro (STC-1 cells) and in native I-cells also showed an important role for the CaSR in the effects of aromatic amino acids on CCK secretion.48–50 Protein hydrolysates stimulate the release of CCK via GPR92 in STC-cells.51
Glucose-induced CCK secretion in humans, although perhaps not a major phenomenon, was unaffected by the sweet taste receptor inhibitor lactisole.54 Immunolocalisation studies confirmed that the sweet taste receptor-specific subunit, TAS1R2, is not coexpressed with CCK.47
GLP-1 and glucose-dependent insulinotropic polypeptide (GIP) are incretin hormones that augment insulin secretion after ingestion of a meal. They are secreted from L-cells and K-cells, respectively. The mechanisms by which L-cells couple glucose detection to GLP-1 secretion have been controversial, but some clarity has begun to emerge (figure 3). Several lines of evidence support that the sweet taste receptor, TAS1R2-TAS1R3, and the taste G-protein, α-gustducin, are required for glucose-stimulated GLP-1 secretion from the small intestine. First, these three proteins are present in human and rodent L-cells of the small intestine.55–57 Second, several sweet taste receptor agonists — sucrose, glucose, fructose and sucralose — can elicit GLP-1 secretion from mouse jejunal and ileal explants, and/or mouse (GLUTag) and human (NCI-H716) enteroendocrine cell lines.56 ,58 ,59 Third, glucose-stimulated GLP-1 secretion is nearly or fully abolished in α-gustducin−/− and TAS1R2−/− and TAS1R3−/− mice.56 ,58 Fourth, sweet taste receptor inhibitors, or RNA silencing of α-gustducin, reduce sucralose-stimulated GLP-1 secretion from GLUTag and NCI-H716 cells.56 ,59 Finally, the sweet taste receptor inhibitor, lactisole, reduced systemic levels of GLP-1 and PYY after intragastric or intraduodenal administration of glucose in humans.54 ,55
Several studies also indicated an important role for the Na+-glucose cotransporter SGLT1 and the ATP-sensitive K+ (KATP) channel. For example, the KATP channel blocker tolbutamide enhances GLP-1 secretion from GLUTag cells or from isolated L-cells from upper small intestine or colon.60 ,61 Furthermore, immunoreactivity for one KATP channel subunit, Kir6.2, has been localised to L-cells.62 By contrast, a related KATP channel blocker, glibenclamide, had no effect on GLP-1 secretion from mouse ileum explants, but did induce GLP-1 secretion from colonic explants, suggesting that this channel is involved in GLP-1 secretion from L-cells of the large, but not small intestine.58 The evidence for a role of SGLT1 is more intriguing, as SGLT1−/− mice exhibit reduced GLP-1 and GIP levels after glucose gavage.63 A role for SGLT1 in glucose-induced GLP-1 release has been confirmed in GLUTag cells and primary L-cells.61 ,64 ,65
A physiological relationship exists between L-cells and enterocytes. SGLT1 and another glucose transporter, GLUT2, are rapidly upregulated in enterocytes in response to the presence of dietary sugars or sweeteners. However, their expression and/or translocation is downregulated in TAS1R3−/− and α-gustducin−/− mice, suggesting a strong correlation between taste receptor activation in L-cells and the modulation of glucose transporter expression in enterocytes, likely through the actions of GLP-2.59 ,66 ,67 Cats which cannot taste sugars do not upregulate SGLT1 expression in response to increased dietary carbohydrate levels.68 ,69
It is unclear whether FFAR1 or GPR120 play the more important role in L-cell and K-cell fatty acid sensing. Both receptors are expressed in L-cells.44 ,70 Oral administration of α-linolenic acid promoted GLP-1 secretion in vivo.70 In STC-1 cells, however, siRNA against GPR120, but not FFAR1, impaired α-linolenic acid-induced GLP-1 secretion. Nevertheless, the secretion of GIP and GLP-1 in response to acute, oral fat diet administration was reduced in FFAR1−/− mice and was associated with a concomitant reduction in insulin secretion and glucose clearance.44
Bitter ligands stimulate GLP-1 secretion from TAS2R-expressing enteroendocrine cell lines (STC-1 and NCI-H716), suggesting that bitter taste receptors could have an impact on glucose and insulin regulation.52 ,71 In fact, in an Amish Family Diabetes Study, a functionally comprised TAS2R was associated with disturbed glucose homeostasis.71
GPRC6A was recently shown to act as an amino acid sensor in GLUTag that enhances GLP-1 secretion.72
Nutrient sensing in the colon
The main function of the colon is to store food residues, secrete mucus and absorb remaining water and electrolytes from the food residue before defecation. The colon exhibits segmental movements (haustrations) and phasic propulsive contractions. The intestinal flora performs fermentation reactions that produce SCFA and flatus.
SCFAs (propionate, acetate and butyrate) are sensed in the lumen by the free fatty acid receptors, FFAR2 and FFAR3.73 In the human colon, FFAR2 immunoreactive cells were colocalised with PYY-containing L-cells but not with 5-HT containing endocrine cells.74 In the rat colon, 5-HT-containing mucosal mast cells were immunoreactive for FFAR2.75 FFAR3-positive cells in the human colon were less abundantly present than FFAR2 cells and contained PYY but not 5-HT or FFAR2.76 α-gustducin Is colocalised with several fatty acid receptors present on endocrine cells in the mouse colon and is required for SCFA-mediated GLP-1 release from mucosal explants.77 The effect of SCFA on the release of GLP-1 was also confirmed in isolated primary L-cells from mouse colon. In FFAR2−/− and FFAR3−/− mice reduced SCFA-induced GLP-1 secretion was observed in vitro and in vivo together with a parallel impairment of glucose tolerance.78
The functional role of TAS1R3 in L-cells of the human colon is not clear.55 GLP-1-secreting cells of the colon and rectum are normally not responsive to sugars, suggesting no role for the sweet taste receptor in hindgut GLP-1 secretion. Even when these cells become glucose-responsive in association with carbohydrate malabsorption, for instance after gastric bypass surgery, this glucose sensitivity is sweet taste receptor independent (the identity of the glucose sensor is unknown).
Contributions of taste receptors to pathophysiology: possible targets for therapy?
Glucose homeostasis and diabetes
Young et al79 ,80 were the first to compare the intestinal levels of sweet taste receptors in patients with well-controlled type 2 diabetes to those without diabetes. Absolute levels of TAS1R2, TAS1R3, α-gustducin or TRPM5 transcripts in the duodenum of healthy people, or patients with type 2 diabetes, were unaffected by acute variations in glycaemia during fasting. However, the intestinal sweet taste receptor system was highly responsive to changes in luminal glucose. During euglycemia, intraduodenal glucose infusion increased TAS1R2 transcript levels in both groups but during hyperglycaemia TAS1R2 mRNA expression decreased in healthy volunteers but not in diabetics. Levels of TAS1R3 did not change significantly. The TAS1R2 dysregulation in patients with type 2 diabetes may potentially increase the risk of postprandial hyperglycaemia since they exhibited increased glucose absorption during hyperglycaemia, as evidenced by an increase in the glucose absorption marker 3-O-methyl-glucose (3-OMG), compared with healthy subjects. As SGLT1 is responsible for the active transport of luminal 3-OMG, these studies confirm that TAS1R2 regulates glucose absorption via SGLT1. The positive association found between luminal glucose-induced changes in some sweet taste receptor transcripts and the secretion of GLP-1 and GIP underscores that sweet taste receptors do have a regulatory role in the release of incretins. Although these findings are supported by in vitro studies, in vivo studies either in rodents or in humans failed to show an effect of artificial sweeteners on the release of satiation hormones, thereby questioning the role of sweet taste receptors in incretin release.56 ,58 ,81–83 Nevertheless, chronic exposure of high-fat fed diabetic mice with the artificial sweetener, oligofructose, increased GLP-1 levels. Thus, the relative importance of glucose transporters versus sweet taste receptors in the release of incretins remains a matter of debate.
Glucose and insulin levels are remarkably normal in TAS1R2−/− and TAS1R3−/− mice, despite the presence of sweet taste ageusia in both genotypes. However, after an oral glucose challenge TAS1R3−/− mice, but not TAS1R2−/− mice, showed increased blood glucose and decreased insulin levels. The TAS1R3−/− phenotype was less severe after an intraperitoneal injection of glucose, indicating that some, but not all, of the glucose dysregulation results from the disruption of normal gut physiology. These findings are somewhat at variance with the studies in patients with type 2 diabetes where a dysregulation of TAS1R2 but not of TAS1R3 was observed during a luminal glucose challenge and resulted in an increased glucose absorption.
It is clear that the regulation of sweet taste receptors and glucose transporters during different glycemic conditions is complex. More studies using different models and conditions are therefore warranted.
Gastric bypass surgery is one of the most effective methods for inducing a marked and sustained weight loss in morbidly obese patients. This procedure also results in amelioration or even complete remission of type 2 diabetes mellitus independent from weight loss.84 The mechanisms involved are incompletely understood but involve restriction, malabsorption and humoral changes. Indeed, bypass surgery decreases the release of the orexigenic hormone, ghrelin, and increases the release of satiation hormones PYY and GLP-1. After gastric bypass surgery, the contact of nutrients with much of the stomach, duodenum and part of the jejunum is bypassed. It can be hypothesised that nutrient sensors on the ghrelin cell are isolated from contact with nutrients thereby probably affecting ghrelin release and, consequently, energy and glucose homeostasis. Additionally, the rapid delivery of undigested nutrients to the lower small intestine may affect the regulation of taste receptors and/or glucose transporters on L-cells, resulting in enhanced release of PYY and GLP-1. Indeed, duodenal–jejunal bypass surgery in a rat model of type 2 diabetes decreased SGLT1 mRNA and SGLT1-mediated transport in jejunum distal to the duodenojejunostomy and improved diabetes independent from body weight loss.85 A similar observation was made in non-obese mice after duodenojejunal bypass surgery.86 Sweet taste receptors regulate SGLT1 and a significant decrease in TAS1R2 and TAS1R3 protein levels in the alimentary limb after gastric bypass have been demonstrated in rats.87 These results suggest that TAS1R3 antagonists or SGLT1 inhibitors may represent a promising target for the treatment of obesity. In fact, LX4211, a dual intestinal SGLT1/renal SGLT2 inhibitor increased GLP-1 levels and PYY levels in patients with type 2 diabetes and improved blood glucose levels and, therefore, mimicked the effects of gastric bypass surgery.88 These results are inconsistent with studies showing that SGLT1 is required for glucose-mediated GLP-1 release by L-cells in vitro.61 ,63 However, due to the carbohydrate malabsorption induced by SGLT1 inhibition, L-cells in the colon also become glucose sensitive and increase GLP-1 levels.89 The authors suggested that this may be mediated by SCFAs produced by colonic fermentation of unabsorbed glucose that interact with FFARs on L-cells to induce GLP-1 and PYY release. A KATP channel-dependent mechanism involving an unknown glucose sensor has also been suggested in glucose-induced hindgut GLP-1 secretion during carbohydrate malabsorption.58
After gastric bypass surgery, patients show a decreased preference for sweet and fatty foods, but the factors involved are unclear. The decision of what to eat is modulated by taste, olfaction and oral textural perception. Taste, in particular, has an important input into food preference because it has direct effects on brain reward circuits that drive eating. It has been suggested that the acuity for sweet taste increases after gastric bypass, potentially leading to increased intensity of perception.90 It remains to be investigated whether endocrine cells in the gut also show changes in sweet taste sensitivity after gastric bypass surgery. Roux en-Y gastric bypass (RYGB) could further reset the food reward system through changes in the release of gut hormones, known to modify activity in brain reward systems and dopaminergic signalling, but also known to modulate taste sensitivity in taste buds.91 ,92 Indeed, a recent study showed that obese patients, after RYGB, had lower brain-hedonic responses to food than patients after gastric banding.93 Additionally, aversive conditioning during the early postsurgical period where unpleasant feelings (eg, dumping) associated with particular foods can lead to conditioning of food aversion, may also play a role.
Irritable bowel syndrome (IBS): FODMAPS and gluten sensitivity
Patients with irritable bowel syndrome (IBS) claim that diet plays a major role in triggering GI symptoms.94 The most common approaches to manage food intolerance in IBS include: (1) the low-FODMAP (fermentable, oligosaccharides disaccharides, monosaccharides and polyols) diet, (2) the gluten-free diet, (3) the elimination diet for food chemicals. How foods trigger functional gut symptoms is unclear.
The low-FODMAP diet
Recent trials showed that dietary FODMAPs, especially fructose and fructans, are dietary triggers for symptom generation, and that restricting their level of intake might lead to durable symptom improvement in patients with IBS.95 ,96 Two mechanisms of action have been put forward for the FODMAP diet: (1) FODMAPs are poorly absorbed in the small intestine and are osmotically active, thereby drawing fluid through the large bowel leading to diarrhoea97; (2) fermentation of the FODMAPs in the colon will generate gases that may be incorporated in volatile end-products.98 The increase in fluid and gas components will lead to luminal distension and may, in the presence of visceral hypersensitivity, induce bloating, flatulence, abdominal pain and motility disturbances.
In this review, we speculate that alterations in nutrient-sensing mechanisms may also play a role in symptom generation in patients with IBS.
Altered short-chain fatty acid sensing?
Fermentation of FODMAPs, such as fructans, in the colon will also result in the production of SCFAs, known to affect local ion secretion99–101 and colonic motility.102 ,103 Indeed, intraluminal application of SCFA in the rat proximal colon accelerated colonic transit by inducing the release of 5-HT that stimulates 5-HT3 receptors located on vagal sensory fibres, resulting in muscle contraction via a vagal reflex pathway involving the release of ACh from the myenteric plexus (figure 4).104 ,105 The role of 5-HT in SCFAs-induced contractions was confirmed in contractility studies with smooth muscle strips in vitro.106 In fact, use of 5-HT3 antagonists has been shown to improve symptoms in patients with IBS.107 We hypothesise that FFAR2 receptors on mucosal mast cells mediate the effect of SCFA on 5-HT release, since enterochromaffin cells and smooth muscle cells do not contain FFARs (figure 4).74 ,75 Cremon et al108 showed that 5-HT release from mucosal biopsy specimens of patients with IBS was increased and correlated with mast cell counts and the severity of the abdominal pain. It can be hypothesised that in patients with IBS, alterations in the sensitivity or number of free fatty acid receptors involved in the SCFA-induced release of 5-HT from mast cells, may play an important role.
As already outlined, FFAR2 and FFAR3 sense SCFAs to affect the release of gut peptides. Patients with IBS have increased fasting and postprandial CCK levels.109 Isolated duodenal I-cells are highly enriched in FFAR2 and FFAR3 mRNA transcripts.105 Under normal conditions these cells mainly sense circulating SCFAs. The concept that small intestinal bacterial overgrowth is a major pathogenic mechanism underlying IBS is still a matter of debate.110–112 Nevertheless, in those patients with IBS with bacterial overgrowth, luminal SCFA may also be sensed by the I-cells and increase CCK secretion (figure 4). Infusion of CCK in patients with IBS can lead to excessive intestinal motor activity and reduced pain thresholds.113 ,114
The postprandial PYY response is significantly increased in hypersensitive compared to normosensitive patients with IBS.109 We propose that FFAR2 activation by SCFAs might be an important trigger for the release PYY/GLP-1 from colonic L-cells (figure 4). Indeed, feeding rats a fructo-oligosaccharide-enriched diet selectively induced the proliferation of FFAR2-positive L-cells.115 Additionally, colonic infusion of SCFA stimulated PPY release in rats. Immunoneutralisation of circulating PYY abolished the effect of SCFA on colonic motility while the effect was mimicked by exogenous PYY infusion.116 In FFAR3−/− mice, characterised by a decrease in PYY expression, intestinal transit rate was increased.117
In conclusion, the low FODMAP diet might also improve symptoms by reducing the levels of SCFAs that act via FFAR2, FFAR3 and transporters to affect ion secretion and GI motility by releasing hormones and neurotransmitters. Studies are warranted to investigate whether alterations in the number or sensitivity of these FFARs occur in patients with IBS. If so, FFAR2 and FFAR3 antagonists, currently under development, may therefore become promising tools for the treatment of IBS.118 Since SCFAs induce neutrophil chemotaxis through FFAR2, a low FODMAP diet or FFAR2 antagonists may, therefore, also help to control inflammation in subsets of patients with IBS with low-grade inflammation.119
Altered carbohydrate sensing?
Effects observed in TAS1R3−/− and GLUT5−/− mice mimic the effects observed in patients with IBS. Due to duodenal carbohydrate malabsorption, both genotypes display a distended proximal colon with the development of gas pockets as a result of subsequent fermentation by intestinal microflora.58 ,120 Additionally, the increased carbohydrate content in the colon of TAS1R3−/− mice, induced robust glucose-induced GLP-1 secretion from the colon which may induce alterations in transit (figure 5). Fructose interacts with TAS1R3 and GLUT5.58 ,120 We hypothesise that the fructose malabsorption in some patients with IBS may be due to alterations in the basal expression of TAS1R3 and GLUT5 or in their regulation in response to dietary fructose. Therefore, future studies are warranted to investigate the mechanisms of altered glucose sensing in patients with IBS.
The gluten-free diet
Gluten is a well known trigger for GI symptoms in the setting of coeliac disease. The existence of gluten intolerance was demonstrated in patients with IBS without celiac disease.121 However, the same authors recently reported that gluten might not be a specific trigger of functional gut symptoms once dietary FODMAPs are reduced.122 Another controlled trial of gluten-free diet showed that gluten alters small bowel permeability in diarrhoea-predominant patients with IBS, particularly in HLA-DQ2/8-positive patients.123 This may result in greater fluid flux toward the lumen and may elicit immune responses that affect afferent nerves resulting in hypersensitivity.
Altered glutamate sensing?
Glutamic acid (Glu) and proline together account for one half or more of the peptide-bound amino acids in gluten. Ingestion of a protein diet rich in L-Glu does not lead to appreciable changes in plasma glutamate concentrations, since glutamate is extensively metabolised by enterocytes and does not reach the portal vein.124 Thus, once proteins such as wheat are digested, dietary Glu may stimulate Glu sensors (e.g TAS1R -TAS1R3, CaSR and metabotropic glutamate receptors (mGluR)) in the stomach and intestine producing local effects on gut function. For example, in dogs, intragastric, but not intraduodenal, administration of monosodium glutamate (MSG) stimulated gastric emptying and induced phasic non-propagating contractions in the upper gut which were blocked by vagotomy.125 Contradictory results have been reported concerning the effect of MSG on gastric emptying in humans.126 ,127
Among 20 natural amino acids investigated, only L-Glu evoked firing of afferent fibres of the vagal gastric branch in rats. The effect is mediated via metabotropic L-Glu receptors that induce the release of 5-HT from mucosal cells interacting with 5-HT3 receptors on afferent fibres.128 Additionally, intragastric infusion of Glu activated brain regions that are directly or indirectly targeted by these vagal inputs.129 Nevertheless, the effect of glutamate on vagal afferent firing seems not to be straightforward, since inhibitory effects of exogenous and endogenous glutamate have been reported on vagal afferent mechanosensitivity involving group II (mGluR2 and 3) and group III mGluR (mGlu R4, 6, 7, 8).130
In view of the beneficial effects that are observed with a gluten-free diet in patients with IBS, we hypothesise that alterations in luminal glutamate sensing may occur that affect vagal nerve activity resulting in altered brain–gut interactions and symptom generation.
The elimination diet for food chemicals
The elimination diet involves restriction of common food allergens, specific chemical substances in foods or medications that contain these chemicals.
Coffee is commonly reported as a trigger for symptoms in patients with IBS.131 It is not clear whether salicylates or caffeine in coffee are the triggers but both components taste bitter and could alter gut function via TAS2Rs.
According to the Rome III classification, functional dyspepsia can be divided into two categories: postprandial distress syndrome and epigastric pain syndrome, suggesting that at least in some patients the disorder is related to food ingestion. While most of the patients report that their symptoms are triggered within 30 min after meal ingestion, few studies have been performed to evaluate the role of specific foods.132 Particularly, meals containing fat seem to induce or exacerbate symptoms.133 Pilichiewicz et al134 reported that ingestion of a high-fat meal, but not of a high-carbohydrate meal, was associated with a substantially greater increase in nausea and pain immediately after completion of ingestion. Furthermore the scores for nausea and pain were related directly to the increased plasma CCK concentrations in patients with functional dyspepsia. It is known that exogenous CCK administration can mimic dyspeptic symptoms.135 It is therefore conceivable that similar to patients with IBS, who frequently overlap with functional dyspepsia patients, alterations in the expression or sensitivity of free fatty acid receptors on endocrine cells may render them more sensitive to dietary fat and exacerbate symptoms.136 Diet-intervention studies, similar to what has been performed in patients with IBS, are warranted to investigate the role of nutrients and, hence, of taste receptors in symptom generation in patients with functional dyspepsia.
Conclusions and perspectives
The observation, and now compelling evidence, for the presence and function of taste receptors in extraoral tissues offers exciting new possibilities for targeted therapeutics in the battle against diseases of and involving the GI tract. Several taste receptor families involved in nutrient sensing in the gut offer potential advantages as drug targets.
Sweet taste receptors
This long history of sweetener chemistry suggests that new compounds of exceptional specificity and high efficacy could be designed. The beneficial effects of artificial sweeteners in the control of body weight and glucose homeostasis are still a matter of debate, but advances in our understanding of the sweet taste receptor biology may therefore help to design new artificial sweeteners with favourable effects.137 Indeed, in contrast with in vitro studies, in vivo studies either in rodents or in humans failed to show an effect of artificial sweeteners on the release of satiation hormones. 56 ,58 ,81–83 Advances in our understanding of the sweet taste receptor biology may therefore help to design new artificial sweeteners with favourable effects. However, their effectiveness may be complicated by the complex interaction between sweet taste receptors and glucose transporters that cross-regulate each other's expression. Additionally, the evidence provided for altered sweet taste receptor control in type 2 diabetes may open new opportunities for drugs that interfere with sweet taste receptor regulation.
In gastric bypass patients, additional studies should help to provide insight to what the effect is of shunting nutrients to more distal regions of the gut in the control of taste receptor expression/glucose transporter regulation, and the release of hormones involved in the regulation of energy and glucose homeostasis. It is tempting to speculate that targeting these receptors/transporters may mimic the effects of gastric bypass surgery in a non-surgical manner. In fact, ongoing trials with SGLT1 inhibitors already show promising results.88
Bitter taste receptors
Evidence suggests that bitter agonists could be considered as good targets to reduce hunger and motility.18 ,22 ,32 However, the design of new bitter drugs may be complicated by the fact that in humans, 25 receptor subtypes exist, each with different ligand selectivities and a different distribution pattern. Nevertheless, there are also some opportunities. Thousands of plant-derived bitter tastants and metabolites are available that could be tested in appropriate physiological models to select those with favourable therapeutic profiles.
Amino acid and free fatty acid receptors
Recent evidence suggests that food choice (low FODMAPs, gluten-free diet) is a key management strategy for the treatment of symptoms in patients with IBS. These studies underscore claims made for many years by alternative practitioners. The mechanisms involved are not completely understood, but indirect evidence suggests that altered carbohydrate sensing, exacerbated sensing of SCFAs by FFARs or of Glu by amino acid receptors may be involved. Determining alterations in basal expression levels of these receptors or in their regulation in response to a diet may help to provide a rationale for the use of FFAR antagonists or Glu receptor antagonists in these patients and may help to predict whether a patient will be responsive to the diet or not. Similar strategies may be useful in patients with functional dyspepsia.
It is clear that the existence and functional role of taste receptors in the gut is a new and fascinating field of research that may lead to the development of new therapeutic drugs that may benefit from the progress made in our understanding of taste receptor function in the tongue.
The author wishes to thank S. Munger for providing helpful discussion.
Competing interests None.
Provenance and peer review Commissioned; externally peer reviewed.