Transient receptor potential (TRP) channels as drug targets for diseases of the digestive system

1The selection of useful food and its effective digestion are essential for survival. To accomplish this goal, the digestive tract possesses a large array of molecular sensors and transducers to make the right choice and guide the process of digestion by appropriate feed-forward and feed-back mechanisms. The physiological task of the alimentary canal is highly challenging because, depending on the nature of ingested food, the gut needs to carry out programs with opposite aims: digestion of food and absorption of nutrients and water, on the one hand, and the elimination of toxins, antigens, pathogens and indigestible waste, on the other hand. Research in the past decade has revealed an elaborate system of molecular sensors serving taste, analyzing the chemical composition of food, monitoring the motor and secretory activity of the gastrointestinal (GI) tract, and detecting adverse conditions in the lumen and wall of the alimentary canal. Transient receptor potential (TRP) channels, so named after the role they play in Drosophila phototransduction, take a prominent position among these sensors and transducers in the digestive system. A number of their implications in GI physiology and pathology were discovered subsequently to the analysis of the mechanism of action of pungent spices.

2The search for and the intake of nutrients are governed by the appearance, smell and taste of food. There is anthropological evidence that the adoption of cooking food has been a key feature in human evolution. Along with cooking, mankind also learned that the taste of food can be heightened by seasoning, and food prepared with the right mixture of spices is one of the major human pleasures. The chemicals responsible for the gustatory and olfactory pleasures of spices are secondary metabolites of plants. Recognition of the chemical qualities of spices must have driven the co-evolution of TRP channels as appropriate sensors in the animal kingdom. The functional implications of TRP channels are, in the first place, congruent with the strategy of plants to discourage predators by the unusual sensory quality of spices. However, by perversion of this principle, humans have learned to enjoy low doses of those deterrent chemicals. This argument is supported by the fact that capsaicin is pungent for mammals, but not birds which are supposed to help distributing the seeds of red pepper, given that the avian orthologue of the TRP vanilloid-1 channel (TRPV1) lacks the binding site for capsaicin.

3The implication of TRP channels in the sensation of deterrent chemicals was first heralded when, in 1997, TRPV1 (then named the vanilloid receptor-1) was identified as the sensor for capsaicin at the genetic and functional level. Responsible for the piquancy of red pepper (Capsicum spp.), the vanilloid capsaicin in its pure form is one of the most painful chemicals we know, yet is also widely used for food seasoning. TRPV1 was soon joined by other TRP channels with unique transduction properties relevant to chemo-, thermo- and mechanosensation as well as sweet, bitter, sour, salt and umami taste sensation. In addition, these molecular sensors can detect specific chemical entities including unpleasant and/or painful toxins, whereby they subserve chemesthesis, the chemical sense distinct from taste and smell.

4Currently, some 30 different TRP subunit genes have been identified in mammals, these subunits being grouped in 6 families. Members of 5 subunit families, specifically of the vanilloid TRP (TRPV), melastatin TRP (TRPM), ankyrin TRP (TRPA), polycystin TRP (TRPP) and canonical or classical TRP (TRPC) family, are relevant to spice sensing, chemo-, thermo- and/or mechanosensation (Table 1). The TRP channel subunits are made of 6 transmembrane domains with a pore between transmembrane domains 5 and 6. Functional TRP channels, thought to be composed as tetramers of four subunits, are opened or closed by conformational changes in the channel protein. TRP channels are only weakly sensitive to depolarization but open in response to changes in temperature, binding of ligands or other alterations of the channel protein. The ion selectivity differs markedly among the family of TRP channels, most of them being non-selective cation channels.

5The objective of this article is to review the occurrence and function of TRP channels in the digestive tract, with a particular emphasis on their roles in health and disease. To keep the wealth of the available information in perspective, overview panels guide the reader to keep track of the functional implications of TRP channels in the alimentary canal (Figs. 1 and 2) and their associations with disorders of the digestive system (Fig. 3). Alterations in their expression and function in disease have raised enormous interest in TRP channels being attractive targets for new therapeutic strategies in gastroenterology. Accordingly, several TRP channel ligands have been developed, and it is a further objective of this article to summarize and discuss the pharmacological opportunities which TRP channel ligands offer to gastroenterology (Fig. 4).

6Although various TRP channels, notably TRPV1, TRPV3, TRPA1, TRPM5 and TRPM8, enable humans to enjoy seasoned food, they subserve a much wider spectrum of functions in the alimentary canal. TRP channels survey the environment of the gut for a large variety of chemicals and toxins that either come in with the food, are formed in the GI tissue, are constituents of the digestive juice, are produced by the GI microbiome, or are generated by tissue injury and inflammation (Table 1). In addition, some TRP channels are involved in mechanosensation or can sense a specific range of temperatures. Apart from their sensory surveillance function, TRP channels in the alimentary canal are relevant to the control of membrane potential and excitability of neurons, epithelial cells, muscle cells and interstitial cells of Cajal (ICCs), play a role in Ca2+ and Mg2+ absorption, govern blood flow, pacemaker activity, motor activity, secretory processes and mucosal homeostasis, and impact on the development of GI cancer.

7In reviewing the functional implications of TRP channels in the alimentary canal, 3 distinct roles come to light (Fig. 1). (1) TRP channels operate as molecular sensors (detectors or primary transducers) of chemical and physical stimuli; (2) TRP channels play a role as downstream or secondary transducers (or effectors) of cell activation mediated by G protein-coupled receptors (GPCRs) or ion channel receptors; and (3) TRP channels function as ion transport channels responsible, e.g., for Ca2+ and Mg2+ homeostasis. Within primary afferent neurons, translation of signals detected by TRP channel detectors into effector responses is carried by two distinct mechanisms: (1) by the local release of neuropeptides from the peripheral fibers of TRP-expressing afferent neurons, which causes changes in local tissue function; and (2) by transmission to the central nervous system resulting in autonomic reflex responses and sensation.

8The existence of TRPV1, the specific sensor for capsaicin, has been envisaged more than 30 years ago, given that most pharmacological effects of capsaicin have been attributed to a specific action on capsaicin-sensitive primary afferent neurons. The role of capsaicin-sensitive afferent neurons and, by indirect inference, the distribution of capsaicin receptors have been traced with two experimental paradigms: stimulation of afferent neurons by acute application of capsaicin, and pretreatment of animals or tissues with high doses of capsaicin (“capsaicin desensitization”) to defunctionalize afferent neurons via excess influx of Ca2+ following stimulation of TRPV1. Capsaicin-sensitive nerve cells may even degenerate if, e.g., a high dose of capsaicin is administered to newborn rats or micromolar concentrations of capsaicin are added to dorsal root ganglion (DRG) cultures. Since the whole sensory neuron is defunctionalized by this approach, referral to a capsaicin-sensitive structure or function in its historical connotation does not necessarily point to an implication of TRPV1. In addition, capsaicin at higher concentrations can act on ion channels unrelated to TRPV1 and on cellular structures lacking TRPV1 such as ICCs. Treatment of adult rats with a neurotoxic dose of capsaicin reduces the number of nodose ganglion neurons that express TRPV1 and project to the stomach as observed 10 days post-treatment. By 67 days post-treatment, much of the vagal afferent innervation of the stomach is restored, mainly because of sprouting of capsaicin-insensitive nerve fibers.

9In the alimentary canal, TRPV1 occurs in extrinsic sensory neurons, intrinsic enteric neurons as well as epithelial and endocrine cells. The major cellular sources of TRPV1 in the rat, guinea-pig and murine alimentary canal are spinal and vagal primary afferent neurons. Both in the DRG and nodose ganglion cells, TRPV1 is restricted to small and medium-sized somata that are known to give rise to unmyelinated (C-) and thinly myelinated (Aδ-) fibers. There are distinct regional differences in the relative proportion of sensory neurons that stain positive for TRPV1. Thus, TRPV1-immunoreactive fibers are considerably more prevalent in visceral than in somatic afferent neurons. In fact, the majority of nodose ganglion neurons projecting to the rodent stomach and of DRG neurons projecting to the rodent gut express TRPV1, while only 32% of the vagal afferent neurons supplying the murine jejunum contain TRPV1.

10Within the rodent and human alimentary canal, TRPV1-positive nerve fibers occur in the musculature, enteric nerve plexuses, arterioles and mucosa. In addition, TRPV1 is present in nerve fibers supplying the taste papillae of the tongue.

11There are regional and species differences in the chemical coding of primary afferent neurons expressing TRPV1. Thus, calcitonin gene-related peptide (CGRP), substance P, somatostatin and other neuropeptides are messenger molecules characteristic of capsaicin-sensitive DRG neurons, while CGRP is absent from vagal afferent neurons containing TRPV1. DRG neurons can be largely differentiated by their binding of isolectin B4 and their responsiveness to different neurotrophins. In adult rodents, the isolectin B4-negative cell population responds to nerve growth factor, while isolectin B4-positive cells respond to the glial cell line-derived family of neurotrophins. However, there is no clear distinction between these groups of DRG neurons in their expression of TRPV1 and the neuropeptides substance P, CGRP and somatostatin. In the rat, TRPV1 is found in both populations of DRG neurons but is more prevalent in isolectin B4-positive cells, whereas in the mouse TRPV1 is largely absent from isolectin B4-positive DRG cells. In both rat and mouse, however, TRPV1 abounds in visceral DRG neurons that bind little isolectin B4 but are rich in CGRP and substance P.

12The presence of TRPV1 in the enteric nervous system has been less extensively studied than that in primary afferent neurons and has remained a matter of controversy. While some authors have localized TRPV1 immunoreactivity to somata of the guinea-pig, porcine and human enteric nervous system, other authors have failed to confirm this location. Importantly, the chemical coding of TRPV1-positive DRG neurons innervating the murine gut and of TRPV1-positive nerve fibers in the murine colon and rectum is distinct from that of enteric neurons. Furthermore, TRPV1 messenger ribonucleic acid (mRNA) disappears from the rat stomach following extrinsic denervation. Enteric neurons that issue projections from the rat colon to the spinal cord are likewise TRPV1-negative. The inconsistencies in the localization of TRPV1 to enteric neurons could in part be explained by the existence of several splice variants of TRPV1 that may substantially differ in their immunoreactivity. For instance, two splicing products of the murine TRPV1 gene have been identified in the DRGs and several tissues including the gut. One of them, TRPV1-α, is equivalent to the rat and human TRPV1, whereas the other one, TRPV1-β, encodes a dominant negative subunit of the TRPV1 channel.

13In addition to its prominent location in neurons, TRPV1 has been localized to non-neural cells of the alimentary canal. Thus, TRPV1-like immunoreactivity is found in the human submandibular gland where it occurs in serous acinar and ductal cells, but not in mucous acinar cells. Furthermore, TRPV1-like immunoreactivity has been localized to gastrin and parietal cells of the stomach as well as to epithelial cells of the esophageal, gastric and small intestinal mucosa of several species including man. With respect to the presence of TRPV1 in these tissues a word of caution is in place, given that TRPV1 antisera can yield false positive staining in non-neural tissues, which persists in TRPV1 knockout mice.

14TRPV1 is a sensor that is stimulated by several spices, noxious heat, and a number of endogenous stimuli, some of which are particularly relevant to the GI tract (Table 1). Among the spices acting on TRPV1, capsaicin (red pepper), piperine (black pepper), gingerol and zingerone (ginger), pungent compounds from onion and garlic, eugenol (clove) and camphor figure most prominently. TRPV1 is also activated by allyl isothiocyanate present in mustard, horseradish and wasabi, resiniferatoxin, a toxin of tropical Euphorbia plants, and “vanillotoxins” present in the venoms of the Trinidad Chevron tarantula and Earth Tiger tarantula. A taste variant of TRPV1 is thought to be involved in nonspecific salt taste perception. Furthermore, TRPV1 contributes to the aversive bitter aftertaste sensation evoked by high concentrations of artificial sweeteners such as saccharin and acesulfame-K and to the metallic taste sensation evoked by high concentrations of CuSO4, ZnSO4 and FeSO4. The burning sensation evoked by ethanol in the mouth and throat is also most likely mediated by TRPV1. The endogenous stimuli that TRPV1 responds to include acid (pH < 6), ammonia (pH > 8), anandamide and other arachidonic acid-derived metabolites such as N-arachidonoyl-dopamine, N-oleoyldopamine, oleoylethanolamide, 12-(S)-hydroperoxy eicosatetraenoic acid, 15-(S)-hydroperoxy eicosatetraenoic acid and leukotriene B4.

15Several lines of evidence indicate that TRPV1 is involved in pain and hyperalgesia throughout the alimentary canal. In addition, TRPV1-expressing vagal afferents in the esophagus may contribute to the cough evoked by gastro-esophageal reflux of gastric contents. In keeping with its nociceptive role, activation of TRPV1 on primary afferent neurons innervating the gut elicits pain in humans and pain-related behavior in rodents (Table 2). Intraesophageal instillation of capsaicin in human volunteers evokes symptoms of heartburn, while the sensitivity to balloon distension or acid exposure remains unchanged. Intragastric administration of capsaicin in humans increases the sensitivity to proximal gastric distension, and ingestion of capsaicin capsules induces gastric sensations of pressure, heartburn and warmth. Infusion of capsaicin into the proximal small intestine of human volunteers likewise evokes sensations of pain, cramps, pressure, warmth and nausea. The capsaicin-induced sensations are most intense in the duodenum and appear to decrease along the intestine, although local capsaicin instillation in ileostomy and colostomy patients is also painful. Ingestion of capsaicin for 7 days leads to a decrease in jejunal mechanosensation and to an increase in the jejunal sensitivity to capsaicin, and oral chili intake for 3 days has been reported to increase rectal sensitivity to urgency. In contrast, ingestion of capsaicin (0.25 mg) capsules three times per day for 4 weeks leads to desensitization of the duodenum to both acute capsaicin administration and balloon distension.

16TRPV1 can be sensitized by many proalgesic factors, and there is increasing evidence that GI inflammation causes chemical and mechanical hyperalgesia at least in part by upregulation and sensitization of TRPV1 (Table 2). Neurotrophins, various inflammatory mediators as well as endocrine factors of the GI mucosa are thought to be involved in this process. Thus, the ability of 5-hydroxytryptamine (5-HT) to sensitize GI afferent neurons to heat, acid and capsaicin is absent in TRPV1 knockout mice. The effect of 5-HT to sensitize colonic DRG neurons is mediated by metabotropic 5-HT2 and 5-HT4 receptors which appear to enhance TRPV1 activity by downstream phosphorylation pathways. Vice versa, depletion of 5-HT from the colon by pretreatment of rats with p-chlorophenylalanine reduces the excitability of DRG neurons by capsaicin and elevates the threshold of distension-induced pain behavior. Referred hyperalgesia associated with caerulein-induced pancreatitis in mice depends on both protease-activated receptor-2 (PAR-2) and TRPV1, TRPV1 being a downstream transducer of the pronociceptive action of PAR-2 stimulation.

17TRPV1 is upregulated not only in inflammation, but also in the absence of overt inflammation (Table 2) as is typical of functional GI disorders. This is true for patients with irritable bowel syndrome in which the increased density of TRPV1 in the rectosigmoid colon correlates with pain severity. A similar correlation between pain intensity and number of mucosal TRPV1-positive nerve fibers is found in patients with quiescent inflammatory bowel disease who continue to complain of abdominal pain. Non-erosive reflux disease, idiopathic rectal hypersensitivity and fecal urgency are other instances of TRPV1 upregulation in the absence of appreciable inflammation. In addition, patients with functional dyspepsia and diarrhea-predominant irritable bowel syndrome exhibit hypersensitivity to the painful sensations evoked by capsaicin-containing capsules. Experimental findings have likewise shown that TRPV1 has a bearing on post-inflammatory colonic hyperalgesia in rodents, given that upregulation of TRPV1 expression and function persists long after the initial inflammatory insult has subsided.

18Although the participation of TRPV1 in chemical nociception figures most prominently, there is evidence that TRPV1 also contributes to mechanical hyperalgesia in the inflamed gut (Table 2). As shown for pelvic afferents innervating the murine colon, TRPV1 is preferentially expressed by mechanoreceptors that respond to distension with a low frequency of firing, the distension responses of these fibers being sensitized by capsaicin or acidosis. Pelvic afferents innervating the smooth muscle and myenteric plexus of the murine rectum are likewise sensitive to low-threshold distension and capsaicin. Accordingly, capsaicin enhances the visceromotor response to colorectal distension (an indirect measure of abdominal pain), this effect being inhibited by the TRPV1 blocker SB-705,498. Similarly, the hypersensitivity to colorectal distension, which is seen after acute stress exposure of adult rats that have been subjected to maternal separation as neonates, is reversed by TRPV1 blockade. This finding is consistent with a report that chronic exposure of adult rats to water avoidance stress for 10 days causes a significant upregulation of TRPV1 (and TRPA1) in DRG neurons supplying the colon, along with an increase in the abdominal withdrawal reflex to colorectal distension. Furthermore, TRPV1 blockade prevents the development of mechanical hyperalgesia that is evoked by repetitive colorectal distension in rats, but does not cause hypoalgesia. High-threshold splanchnic afferents innervating the mesentery and serosa of the colon fail to be sensitized in rats with dextran sulfate sodium (DSS)-induced colitis, yet the response of these fibers to mechanical probing is attenuated by a TRPV1 blocker in inflammation but not health. The role of TRPV1 in mechanical pain seems to be that of a secondary transducer, because a mechanodetector role of TRPV1 has not yet been substantiated. The implication of TRPV1 in inflammatory hypersensitivity is likewise that of a secondary transducer, TRPV1 channel activity being enhanced by multiple mechanisms triggered by receptors for inflammatory mediators and neurotrophins.

19A prominent role of vagal afferent neurons supplying the upper GI tract is to cause nausea and vomiting if the ingested food is deemed potentially or actually toxic. In view of their broad spectrum of sensory modalities, distinct TRP channels are likely to play a role in these processes. There is indeed experimental evidence that activation of TRPV1 has an impact on emesis. Thus, subcutaneous or intracerebroventricular administration of capsaicin or resiniferatoxin has been reported to cause retching and vomiting in the dog, ferret and Suncus marinus, the house musk shrew. The proemetic action of TRPV1 activation appears to reverse quickly to an antiemetic action, given that vomiting caused by stimulation of the medial solitary nucleus in the brainstem, radiation, systemic administration of cisplatin, loperamide or apomorphine or intragastric administration of CuSO4 is depressed by capsaicin and resiniferatoxin. Since the mechanism behind the dual proemetic and antiemetic action of TRPV1 agonists is not understood, the existence of a novel vanilloid receptor has been envisaged. Non-pungent agonists at TRPV1 such as arvanil, olvanil, anandamide and N-arachidonoyl-dopamine fail to evoke emesis in the ferret but are able to blunt vomiting evoked by morphine-6-glucuronide, apomorphine, cisplatin and CuSO4.

20TRPV1 is a heat sensor, which explains why the sensation caused by capsaicin is described as “hot” and “burning”. Although capsaicin and related TRPV1 agonists evoke a thermoregulatory response, TRPV1 knockout mice have a normal body temperature and do not seem to have a deficit in heat sensing, except that heat hyperalgesia in response to inflammation or heat injury and fever in response to the bacterial pyrogen lipopolysaccharide are attenuated. The maintenance of basal thermoregulation in TRPV1 knockout mice is probably due to developmental compensations, given that many TRPV1 blockers cause substantial hyperthermia in mice, rats, dogs, monkeys and humans. It need be inferred, therefore, that TRPV1 is tonically active to control core body temperature, lowering body temperature through suppression of autonomic cold-defense effectors.

21The hyperthermic action of TRPV1 blockers involves cutaneous vasoconstriction and shivering-related thermogenesis, but not warmth-seeking behavior, and is inhibited by the antipyretic drug acetaminophen. Because the magnitude of the hyperthermic effect of the TRPV1 blocker AMG0347 is independent of the baseline temperature, it has been concluded that the rise of body temperature results from blockade of tonic TRPV1 activation by non-thermal factors. Further analysis suggests that the hyperthermic response to TRPV1 antagonists depends on blockade of the proton mode of TRPV1 activation, either alone or together with the capsaicin mode of TRPV1 activation. However, this issue has not yet been settled because some TRPV1 antagonists that do not block the proton mode of TRPV1 activation can elicit hyperthermia and other TRPV1 blockers that do not cause hyperthermia affect the proton mode of TRPV1 stimulation. It thus awaits to be confirmed that hyperthermia is in fact driven by a low tissue pH, either alone or together with endovanilloid agonists and that the TRPV1-mediated thermoregulatory action takes place in the abdominal cavity in which the stomach and colon have an acidic environment. The increase in metabolic rate and thermogenesis evoked by intrajejunal administration of non-pungent capsaicin analogs (capsinoids) to mice also seems to be mediated by TRPV1 within the GI tract.

22When TRPV1-expressing sensory nerve fibers are activated, they release peptide transmitters from their peripheral endings and in this way modify GI vascular, immune and smooth muscle functions. Following tissue irritation or injury, some of these reactions (e.g., vasodilatation and plasma protein extravasation) contribute to the process of neurogenic inflammation. The TRPV1 agonist capsaicin and mustard oil (a TRPA1 and TRPV1 agonist) have been instrumental in discovering and analyzing this phenomenon. The messengers involved in the efferent-like mode of operation include CGRP, somatostatin and the tachykinins substance P and neurokinin A.

23Administration of capsaicin to the esophageal, gastric and intestinal mucosa increases mucosal blood flow, a response which is mimicked by exposure to excess acid. The acid- and carbon dioxide-evoked hyperemia in the esophageal and duodenal mucosa is inhibited by the TRPV1 antagonist capsazepine, which indicates that acid activates TRPV1 on sensory nerve fibers. Through this mechanism, which also includes increases in bicarbonate and mucus secretion, TRPV1-positive sensory nerve fibers are able to protect the esophageal, gastric and intestinal mucosa from a variety of injurious insults. Paradoxically, knockout of TRPV1 has been reported to ameliorate acid-induced injury in the esophagus and stomach. Analysis of this observation in the stomach suggests that disruption of the TRPV1 gene causes a compensatory upregulation of several protective mechanisms in the gastric mucosa. These counterregulatory mechanisms may also explain why acid-induced release of CGRP from the murine stomach remains unchanged in TRPV1 knockout mice.

24Apart from protecting the GI mucosa, TRPV1 activation has been found to exacerbate inflammation and injury in certain models of ileitis, colitis and pancreatitis (Table 2). In addition, capsaicin- as well as acid-evoked stimulation of epithelial TRPV1 in the feline and human esophagus has the potential to give rise to esophagitis. In the murine gastric mucosa, ethanol-induced injury appears to involve TRPV1-mediated release of neuronal substance P and subsequent formation of reactive oxygen species. Experimental colitis in the mouse is attenuated by TRPV1 antagonism or knockout (Table 2). Emerging evidence indicates that TRPV1 contributes to pancreatic islet inflammation associated with type I diabetes and, in addition, plays a role in insulin-dependent glucose regulation, type II diabetes, adipogenesis and obesity. The observation that stimulation of TRPV1, under some conditions, reduces and, under other conditions, exaggerates tissue inflammation and injury may reflect stimulus- and tissue-dependent differences in the process of neurogenic inflammation. In particular, the peptide mediators of TRPV1-positive afferent neurons cover a wide spectrum of actions: Tachykinins facilitate inflammation, while CGRP promotes vasodilatation but not necessarily inflammation, and somatostatin is capable of inhibiting inflammatory processes.

25The messengers released from capsaicin-sensitive afferent nerve fibers in the gut act on enteric nervous system, GI smooth muscle and epithelium to modify motility and secretion. Tachykinins and adenosine triphosphate stimulate motor activity, while CGRP, vasoactive intestinal polypeptide and nitric oxide account for motor inhibition caused by TRPV1 activation. Local motor effects mediated by TRPV1 may become operative when the GI tract is disturbed by endogenous or exogenous irritants. For instance, TRPV1 and substance P are involved in the acid-evoked contraction of opossum esophageal longitudinal muscle. The circular muscle of the isolated human gut is relaxed by capsaicin, and this effect appears to be mediated by nitric oxide and, to some extent, vasoactive intestinal polypeptide. Despite these motor effects of TRPV1 stimulation in vitro, the motor activity of the small and large intestine of experimental animals is maintained at a physiological level following functional ablation of capsaicin-sensitive afferent neurons. Likewise, no obvious changes in GI motor function have thus far been reported to occur in mice deficient in TRPV1.

26Ingestion of capsaicin by humans increases amplitude and velocity of esophageal pressure waves and facilitates secondary peristalsis due to air injection, decreases proximal gastric tone, inhibits phasic contractions of the proximal stomach and inhibits gastric emptying, but does not significantly alter orocecal transit time. The capsaicin-induced improvement of esophageal motility has been observed in healthy volunteers as well as in patients with gastro-esophageal reflux disease. The relevance of TRPV1 to GI motor control may be most pronounced under pathopysiological conditions when stimulation of TRPV1 on sensory nerve fibers causes local messenger release and activation of sympathetic reflexes. In this way, laparotomy or peritoneal irritation slows GI transit and may even cause ileus. This contention is also borne out by the observation that the gastroparesis associated with experimental colitis is relieved by TRPV1 blockers.

27Secretory functions in the alimentary canal are subject to regulation by TRPV1 expressed in neurons as well as epithelial cells. In the human submandibular gland, TRPV1 has been found to activate acinar cells, to stimulate the trafficking of aquaporin-5 to the cell membrane and to cause salivary secretion. Acid- or capsaicin-evoked stimulation of epithelial TRPV1 in the feline and human esophagus induces the formation of platelet-activating factor. Exposure of human isolated antral glands to capsaicin causes release of gastrin and somatostatin, an effect that is thought to be mediated by TRPV1 expressed by gastrin cells, because it is blunted by a chili-rich diet ingested for 3 weeks. Distension of the guinea-pig colon or exposure of the guinea-pig and human colon to H2S in vitro evokes a prosecretory effect which depends on activation of TRPV1-expressing sensory nerve fibers, release of substance P and stimulation of cholinergic secretomotor neurons.

28Upregulation of TRPV1 tissue levels has been observed in a number of inflammatory and functional GI disorders (Table 2). Thus, both erosive and non-erosive reflux diseases are associated with increased levels of TRPV1 in the esophageal mucosa. The upregulation of TRPV1 in the mucosa of patients with erosive esophagitis is associated with elevated expression of nerve growth factor and glial cell line-derived neurotrophic factor. Idiopathic rectal hypersensitivity and fecal urgency and Hirschsprung's disease are instances of TRPV1 upregulation in the absence of inflammation. Patients with uninvestigated dyspepsia have been found hypersensitive to intrajejunal capsaicin infusion, and a proportion of patients with functional dyspepsia are more responsive to ingestion of capsaicin capsules than healthy controls. A role of TRPV1 in upper GI pain is also suggested by the antinociceptive effect which prolonged treatment with capsaicin-containing capsules has in healthy volunteers and patients with functional dyspepsia. Furthermore, the homozygous G315C polymorphism of the TRPV1 gene has been found to be inversely related to symptom severity in functional dyspepsia patients.

29Patients suffering from diarrhea-predominant irritable bowel syndrome exhibit hypersensitivity to the painful and burning sensations which chili-containing capsules and food elicit in the GI tract. The TRPV1-positive innervation of the mucosa in the rectosigmoid colon is increased in patients with inflammatory bowel disease, in patients with irritable bowel syndrome as well as in patients with quiescent inflammatory bowel disease who continue to complain of irritable bowel syndrome-like symptoms. Importantly, the density of TRPV1-positive nerve fibers in the rectosigmoid colon correlates with pain severity in both patients with irritable bowel syndrome and patients with symptomatic but quiescent inflammatory bowel disease.

30Recognition of TRPV1 as a multimodal nocisensor, its sensitization by a number of proinflammatory and proalgesic pathways, its upregulation under conditions of hyperalgesia and its apparent implication in experimental colitis have made this TRP channel an attractive target for novel antinociceptive and antiinflammatory drugs. Pharmacologically, the function of TRPV1 can be manipulated by two principal approaches: stimulant/defunctionalizing TRPV1 agonists and TRPV1 antagonists. The mechanism and result of the two approaches are profoundly different. While TRPV1 antagonists specifically modify the function of the ion channel, stimulant/defunctionalizing TRPV1 agonists target the cellular function of capsaicin-sensitive afferent neurons. The “desensitization” that is brought about by capsaicin, resiniferatoxin or non-pungent TRPV1 ligands reflects “defunctionalization” of the whole afferent neuron expressing TRPV1 for a prolonged period of time.

31At present, there is only evidence for a beneficial effect of stimulant/defunctionalizing TRPV1 agonists in GI disease. Acute ingestion of capsaicin has a protective effect against aspirin-induced erosions in the gastric mucosa of human volunteers. In addition, capsaicin can improve esophageal motility in patients with gastro-esophageal reflux disease. Prolonged intake of capsaicin appears to desensitize afferent nerve fibers against noxious stimulation of the upper GI tract. Thus, ingestion of capsaicin (0.25 mg) capsules by healthy volunteers 3 times per day for 4 weeks blunts the pain response evoked by duodenal capsaicin administration and balloon distension. Similarly, treatment of patients suffering from functional dyspepsia with capsaicin-containing capsules for 5 weeks leads to a significant reduction of pain symptoms. Since there is some information that the prevalence of heartburn symptoms is low in Asian countries in which the intake of chili-rich food is common, it has been proposed that chronic ingestion of capsaicin may be beneficial in patients with gastro-esophageal reflux disease and functional dyspepsia.

32A great deal of effort has been put into developing compounds that block TRPV1 activation in a competitive or noncompetitive manner. Some TRPV1 blockers exhibit species differences in their activity and/or act in a stimulus-specific manner, i.e., differentially inhibit the activation of TRPV1 by capsaicin, heat, and acid. The design of stimulus-specific blockers is of particular importance because a number of TRPV1 antagonists turned out to have a pronounced hyperthermic effect that limits their clinical usefulness. Apart from their thermoregulatory perils, TRPV1 antagonists are liable to impair the perception of potentially injurious heat. Furthermore, blockade of TRPV1 will interfere with the physiological role of this nocicensor in surveying the physical and chemical environment and, if necessary, in initiating protective responses. Such a role is obvious in the GI tract in which capsaicin-sensitive afferent neurons constitute a neural alarm system which helps maintaining mucosal homeostasis in the face of pending injury. Other caveats are posed by the incomplete analysis of the involvement of TRPV1 in GI disease. For instance, the upregulation of TRPV1 in gut disorders does not necessarily reflect a causal implication of TRPV1 but may equally well mirror an epiphenomenon of the disease process.

33There are several opportunities to design TRPV1 blockers such that they can be specifically targeted at aberrantly expressed or aberrantly operating TRPV1 channels while sparing their physiological function. These opportunities include the development of modality-specific TRPV1 blockers, of uncompetitive TRPV1 antagonists, of compounds that selectively address certain TRPV1 splice variants that may be particularly disease-relevant, compounds that prevent TRPV1 sensitization, and compounds that prevent the recruitment and trafficking of TRPV1 to the cell membrane. If these approaches turn out to be successful, TRPV1 blockers may enjoy a wide spectrum of usefulness in GI disease which includes, pain, hyperalgesia and inflammation associated with gastro-esophageal reflux disease, inflammatory bowel disease, functional Gi disorders including functional dyspepsia and irritable bowel syndrome, disturbances of GI motor activity, and nausea and emesis.

34TRPV2 has been localized to rodent DRG and nodose ganglion neurons supplying the GI tract. In contrast to TRPV1, TRPV2 occurs primarily in medium and large diameter afferent neurons. In the rat oral mucosa, TRPV2 is not only present in subepithelial nerve fibers, but also in junctional epithelial cells surrounding each tooth, Langerhans cells, dendritic cells, macrophages and endothelial cells of venules, but not in oral epithelial cells. In the rat and guinea-pig small intestine, TRPV2 is found in nerve fibers within the muscularis and myenteric plexus as well as in myenteric cell bodies. In addition, TRPV2 occurs in the β-cells of murine pancreatic islets and in the insulinoma cell line MIN6. Immunohistochemistry of human tissues has localized TRPV2 to epithelial cells of the salivary glands, pancreatic duct, stomach, duodenum and colon.

35TRPV3 is highly expressed by epithelial cells in the nose and tongue. It likewise occurs in DRG and nodose ganglion neurons, but retrograde labeling has failed to identify TRPV3 in vagal afferent neurons supplying the murine stomach. However, the muscle and mucosa of the murine stomach and small intestine do contain TRPV3 mRNA. In addition, TRPV3 is expressed by superficial epithelial cells of the murine distal colon, but not by superficial epithelial cells of the stomach, duodenum and proximal colon.

36In terms of their thermosensory modalities, TRPV2 and TRPV3 belong to the group of thermo-TRP channels because they are activated by specific temperatures. While TRPV2 is typically stimulated by temperatures above 52 °C, TRPV3 responds to temperatures in the innocuous physiological range. TRPV2 and in particular TRPV3 are sensitive to a variety of chemical entities and—together with TRPA1, TRPM8 and TRPV1—TRPV3 is one of the multimodal spice sensors (Table 1). Noteworthy, TRPV3 is activated by a range of monoterpenoid compounds including carvacrol (oregano), thymol (thyme) and menthol (mint), although these compounds are also recognized by other TRP channels such as TRPA1 and TRPM8.

37The expression of TRPV2 and TRPV3 in the oral cavity, nose and tongue together with their sensory modalities suggests that these TRP channels play a role in probing food for its thermal and chemical qualities. Specific implications of TRPV2 and TRPV3 in the GI tract have not yet been explored. There is evidence, however, that TRPV2 plays a role in the autocrine regulation of pancreatic β-cells, since insulin causes translocation and insertion of TRPV2 into the plasma membrane and the insulin-evoked calcium entry is prevented by TRPV2 inhibition or knockdown. In addition, the release of insulin induced by a high glucose concentration depends on TRPV2.

38A study analyzing a possible association between genetic variability (as examined for 392 single-nucleotide polymorphisms) in 43 fatty acid metabolism-related genes and risk for colorectal cancer in 1225 patients with cancer and 2032 controls has shown that TRPV3 is associated with a higher risk for development of colorectal cancer.

39In the GI tract, TRPV4 occurs primarily in fibers of extrinsic primary afferent neurons, although some epithelial and other cells have also been reported to stain positively for this TRPV channel subunit. TRPV4 is present in the nodose ganglion, DRG, stomach, small intestine and colon of rodents. Retrograde labeling shows that vagal afferent neurons projecting to the murine forestomach contain TRPV4 and that, in these neurons, TRPV4 is coexpressed with TRPV1, TRPV2 and TRPA1 to various degrees. Similarly, DRG neurons projecting to the murine pancreas synthesize both TRPV4 and TRPA1. The expression of TRPV4 in the DRG neurons that project to the murine and human intestine via mesenteric and pelvic nerves exhibits pronounced regional differences. Thus, TRPV4 is particularly abundant in DRG neurons supplying the colon, in which it is coexpressed with CGRP. In the human colon, TRPV4 is found in fine nerve fibers associated with blood vessels in the submucosa and serosa, while the myenteric plexus, the circular and the longitudinal smooth muscles are largely negative.

40In the murine colon, TRPV4 occurs in brush-bordered epithelial cells, but not in mucus-secreting epithelial cells, as well as in mucosal glial cells and unidentified cells of the submucosal and muscular layers. TRPV4 is likewise expressed by epithelial cells of the human colon and by the human colon carcinoma cell line Caco-2.

41A further location of TRPV4 is in the hepatobiliary tract in which it is expressed by ciliated cholangiocytes, as shown for both the rat, murine and human liver and bile system. In these cells, TRPV4 occurs specifically in the apical membrane and the cilia which are thought to subserve a mechano- and osmosensory role.

42The sensory modalities of TRPV4 (Table 1) include strong acidosis, hypotonicity, warmth, mechanical stimuli such as distension of the gut, 5,6-epoxyeicosatrienoic acid (an endogenously formed metabolite of arachidonic acid), and the synthetic phorbol ester 4-α-phorbol 12,13-didecanoate. Although it has been surmised that the mechanosensory function of TRPV4 reflects a secondary transducer role, a patch-clamp study indicates that TRPV4 is a primary detector of mechanical forces.

43TRPV4 has turned out to play a major role in mechanical pain and hyperalgesia in the colon. TRPV4 agonists excite colonic afferent neurons in the mouse and enhance the mechanosensory responses of colonic serosal and mesenteric afferent nerve fibers, while the mechanosensitivity of these high-threshold afferent nerve fibers is substantially attenuated in TRPV4 knockout mice. The TRPV4 agonist-evoked sensitization of colonic afferent nerve fibers to mechanical stimuli is associated with mechanical hyperalgesia, as the visceromotor response to colorectal distension is enhanced. Vice versa, TRPV4 knockout or intravertebral pretreatment of mice with TRPV4-directed small interfering RNA (siRNA) reduces the visceromotor response to colonic distension in the noxious range. Taken these findings together, the specific contribution to colonic high-threshold mechanosensory function makes TRPV4 currently the only nociceptor-specific TRP channel in the gut.

44TRPV4 and PAR-2 are colocalized in spinal afferent neurons, in which these two entities interact with each other to cause mechanical hyperalgesia in the murine colon. Thus, the excitatory effect of TRPV4 stimulation on firing of colonic afferents is increased by a PAR-2 agonist, and a subnociceptive dose of a PAR-2 agonist is able to sensitize colonic afferents to a subnociceptive dose of a TRPV4 agonist. Activation of PAR-2 likewise evokes discharge of action potentials in colonic afferent nerve fibers and enhances the visceromotor response to colorectal distension. These effects of PAR-2 agonism are absent in TRPV4 knockout mice as well as after pretreatment with TRPV4 siRNA.

45Based on these findings it is inferred that TRPV4 plays two distinct roles in colonic pain and hyperalgesia. Under basal conditions, TRPV4 is involved in mechanical nociception, and in this capacity is likely to operate as primary mechanodetector. In addition, TRPV4 mediates the mechanical hyperalgesia evoked by PAR-2 agonism. In this instance, TRPV4 seems to act as secondary transducer of PAR-2, a conclusion that is consistent with the observation that this secondary transducer role of TRPV4 is not confined to PAR-2. Thus, 5-HT and histamine can likewise enhance the expression of TRPV4 in DRG neurons projecting to the murine colon and are able to increase the ability of TRPV4 agonism to induce mechanical hyperalgesia. Importantly, the colonic hypersensitivity to colorectal distension evoked by 5-HT or histamine is attenuated by TRPV4 siRNA.

46In DRG neurons projecting to the murine colon, TRPV4 is not only colocalized with PAR-2 but also with PAR-4. Unlike PAR-2 agonism, activation of PAR-4 counteracts mechanical pain and hyperalgesia. Thus, the excitation of DRG neurons by PAR-2 and TRPV4 agonists is reduced and the ensuing mechanical hypersensitivity to colonic distension is attenuated by a PAR-4 agonist.

47TRPV4 is also likely to contribute to the pain associated with pancreatitis. Intraductal administration of a TRPV4 agonist to the murine pancreas causes expression of c-Fos in the spinal cord. Deletion of the TRPV4 gene inhibits the input to the spinal cord and the pain behavior associated with experimental pancreatitis due to caerulein.

48Emerging evidence attributes TRPV4 a proinflammatory role in the colonic mucosa. Exposure of Caco-2 cells and the human intestinal epithelial cell line T84 to the TRPV4 agonist 4-α-phorbol 12,13-didecanoate increases the intracellular calcium concentration and causes release of chemokines. In keeping with these effects, intraluminal administration of the TRPV4 agonist to the murine colon evokes a transient increase in paracellular permeability and colitis. Furthermore, the expression of TRPV4 in epithelial cells of the colon is significantly enhanced in mice with DSS-induced colitis.

49Hypotonicity causes an increase in the intracellular calcium in cultured murine cholangiocytes and rat intrahepatic bile duct units, a response that is mimicked by a TRPV4 agonist and impaired by TRPV4 knockdown. The hypotonicity-induced activation of cholangiocytes leads to TRPV4-dependent bicarbonate secretion, the main determinant of ductal bile formation. This mechanism operates as well in vivo, since intrabiliary infusion of a TRPV4 agonist increases luminal bicarbonate secretion and bile flow in the rat, this effect being prevented by TRPV4 blockade. In polycystic kidney rats in which mutations in the polycystic kidney and hepatic disease 1 gene cause cholangiocytes to undergo hyperproliferation and to form cysts, TRPV4 is overexpressed and mislocalized in the cholangiocytes, given that TRPV4 is present intracellularly rather than at the apical membrane and the cilia. TRPV4 agonists attenuate the growth of cultured cholangiocytes from polycystic kidney rats and retard the formation of cysts in culture, whereas the growth of cholangiocytes from normal rats is not altered.

50TRPV4 channelopathies are responsible for a number of hereditary sensory and motor neuropathies as well as diseases with defects in bone development. Thus, mutations of TRPV4 are found in congenital distal spinal muscular atrophy and scapuloperoneal spinal muscular atrophy (two hereditary motor neuropathies), hereditary motor and sensory neuropathy of type 2C (also known as Charcot–Marie–Tooth disease type 2C) and other rare motor neuropathies. Diseases with defective bone development such as metatropic dysplasia, brachyolmia, spondylo-epiphyseal dysplasia of Kozlowski type, spondylo-epiphyseal dysplasia of Maroteaux type (pseudo-Morquio syndrome type 2) and parastremmatic dysplasia are also associated with TRPV4 channelopathies.

51Although the GI tract is not the prime victim of these diseases, gastro-esophageal reflux has been observed in hereditary sensory neuropathy of type 1B. Little is known, however, whether the hereditary sensory neuropathies of types 1A, 1B and 1D impact on GI sensation and nociception, although the available reports do not mention any conspicuous GI disease phenotype.

52TRPV4 in sensory neurons is upregulated in inflammatory bowel disease, and serosal blood vessels in tissues with active colitis are more densely innervated by TRPV4-positive nerve fibers than in biopsies from healthy controls. Expression of TRPV4 in colonic epithelial cells of mice with experimental colitis is likewise upregulated, while the levels of TRPV4 mRNA in mucosal biopsies taken from patients with inflammatory bowel disease have been found unaltered. However, the colonic mucosa of inflammatory bowel disease patients is infiltrated by TRPV4-positive CD45-positive leukocytes which are absent from the mucosa of healthy controls.

53In patients with polycystic liver diseases (autosomal-recessive polycystic kidney disease and autosomal-dominant polycystic kidney disease) TRPV4 is overexpressed in cholangiocytes, a finding that is reproduced in polycystic kidney rats, a model for autosomal-recessive polycystic kidney disease.

54Taken together, the experimental findings related to TRPV4 function in the digestive system attribute this TRP channel a pathophysiological role in inflammation, pain, hyperalgesia and polycystic liver disease. This inference if supported by the overexpression of TRPV4 in experimental colitis, inflammatory bowel disease and polycystic liver disease. Since TRPV4 appears to be a common mediator of mechanical hyperalgesia in the colon, it presents itself as a novel therapeutic target in patients with abdominal pain. In addition to their antinociceptive and antihyperalgesic potential, TRPV4 blockers may be useful as antiinflammatory drugs. To the contrary, TRPV4 agonists seem to be beneficial in polycystic liver disease in which they can inhibit cholangiocyte proliferation and cyst formation. Indeed, TRPV4 agonists and blockers are already on the horizon and await to be assessed for their therapeutic efficacy and safety.

55TRPV5 and TRPV6 are expressed in the mucosa of the human and rodent small and large intestine. While TRPV5 occurs primarily in the kidneys, TRPV6 is predominantly expressed in intestinal epithelial cells and has also been localized to human colon cancer (Caco-2) cells. In addition, TRPV5 mRNA occurs in rat nodose ganglion neurons expressing cholecystokinin (CCK) CCK1 receptors, which are likely to project to the GI tract.

56TRPV5 and TRPV6 are the most Ca2+-selective members of the TRP ion channel family, play an important role in intestinal Ca2+ absorption, and undergo calcium-induced inactivation via a phospholipase C-mediated pathway. The functional roles of TRPV5 and TRPV6 are interconnected with each other, given that TRPV5 knockout mice upregulate their intestinal TRPV6 expression to compensate for the negative Ca2+ balance caused by the loss of TRPV5-mediated Ca2+ reabsorption in the kidney. In contrast, TRPV6 knockout mice do not display any compensatory mechanism, which results in secondary hyperparathyroidism.

57The expression and function of both TRPV5 and TRPV6 are subject to regulation by dietary conditions and multiple calciotropic hormones. Calcium deficiency in the diet is an important factor that increases the expression of TRPV6 in the duodenal mucosa of mice. Short chain fatty acids, the fermentation products of fructooligosaccharides, likewise increase expression of TRPV6 and calcium absorption in the rat colorectal epithelium as well as in Caco-2 cells. Curcumin, a polyphenol present in turmeric, is a ligand of the nuclear vitamin D receptor, activation of which enhances the expression of TRPV6 mRNA in Caco-2 cells.

58Parathyroid hormone and 1,25-dihydroxyvitamin D3 are particularly relevant to the expression of TRPV5 and TRPV6 in the intestinal mucosa. In chicken intestinal epithelial cells, 1,25-dihydroxyvitamin D3 and parathyroid hormone increase calcium uptake by stimulating a protein kinase A-mediated pathway to release β-glucuronidase, which in turn activates TRPV6. Consistent with this secondary transducer role of TRPV6 is the finding that the 1,25-dihydroxyvitamin D3-induced calcium uptake is abolished by TRPV6 siRNA. In normal human duodenal mucosa obtained at endoscopy, TRPV6 expression is likewise increased following exposure to 1,25-dihydroxyvitamin D3. Although there is a close functional interconnection between vitamin D and intestinal TRPV6, these factors can also operate independently of each other. Thus, the 1,25-dihydroxyvitamin D3-induced intestinal calcium transport in the murine duodenum does not entirely depend on TRPV6, because intestinal calcium uptake also occurs in TRPV6 knockout mice. In addition, intestinal calcium absorption and duodenal TRPV6 expression are upregulated in pregnant and lactating mice even in the absence of the nuclear vitamin D receptor. Other hormones that impact on intestinal calcium absorption include prolactin which induces TRPV6 mRNA in the rat duodenal epithelium, synergizes with 1,25-dihydroxyvitamin D3 in enhancing TRPV6 expression and stimulates duodenal calcium transport.

59Calcium homeostasis is disturbed in patients with Crohn's disease. Similarly, heterozygous mice carrying a deletion in the adenine- and uridine-rich elements of the tumor necrosis factor gene, considered to be a model for Crohn's disease, display an increase in bone resorption along with a decrease in the duodenal expression of TRPV6 mRNA. TRPV6 in the intestinal epithelium also appears to have a bearing on colon cancer development. Thus, TRPV6 expression is significantly enhanced in the Citrobacter rodentium-induced transmissible murine colonic hyperplasia model. Whereas in the normal colon TRPV6 is restricted to the apical membrane of absorptive enterocytes, in the hyperplasia model TRPV6 is also distributed to the proliferating zone of the colonic crypts, in which it occurs in the basolateral membrane and perinuclear area of many epithelial cells. When the animals are fed with a calcium-rich diet, the overexpression of TRPV6 in the hyperplasia model is reversed.

60Increased levels of colonic TRPV6 are associated with early-stage colon cancer. Thus, TRPV6 has been found to be overexpressed in 66% of stage I tumors and in 17% of stage II tumors but is barely detectable in stages III and IV tumors. These results suggest that aberrant function of TRPV6 is associated with early colon carcinogenesis but not with frank malignancy. This inference is supported by experiments with the human colon carcinoma cell line Caco-2, which constitutively express high levels of TRPV6. Treatment of these cells with TRPV6-specific siRNA significantly reduces TRPV6 mRNA levels by about 50%, which is associated with a 40% reduction of cell proliferation and a more than twofold increase in apoptosis. Together with experiments in mice with colon crypt hyperplasia, these findings indicate that TRPV6 promotes the proliferation of colonic epithelial cells and may be a pathogenetic factor in the early stages of colon cancer development.

61Pharmacological interference with TRPV6 in the digestive system is likely to interfere with calcium absorption and bone mineralization as well as with epithelial cell hyperplasia and early stages of malignancies. Thus, blockade of TRPV6 can be considered as a measure to prevent and inhibit the development of colon cancer, but this effect needs to be balanced against any negative impact on calcium homeostasis.

62In the GI tract, TRPA1 occurs in 3 distinct cellular systems: extrinsic primary afferent neurons, intrinsic enteric neurons and endocrine cells of the mucosa.

63Primary sensory neurons constitute the most significant source of TRPA1 in the body, given that TRPA1 is abundantly expressed by trigeminal ganglion, nodose ganglion and DRG neurons. In rodent and human DRG neurons, especially in those projecting to the gut, TRPA1 is preferentially expressed by small and medium diameter cell bodies in which it is colocalized with TRPV1 and CGRP. Similarly, TRPA1 on nodose ganglion neurons in the guinea-pig is primarily associated with small to medium diameter somata and co-expressed with PAR-2. In the periphery, TRPA1 has been found in the stomach, small intestine, colon and pancreas of the rat and mouse, the major source being peripheral nerve fibers of DRG and nodose ganglion neurons. The distribution of TRPA1 in the rodent gut is similar to that in the canine gut, given that TRPA1 is strongly expressed in the stomach, pancreas, small intestine and colon of the dog.

64The other sources of TRPA1 in the GI tract are neurons of the human and murine enteric nervous system as well as 5-HT-releasing enterochromaffin cells and CCK-releasing endocrine cells of the human, rat and murine GI mucosa. The murine neuroendocrine cell line STC-1, the rat endocrine cell line RIN14B and the human pancreatic endocrine cell line QGP-1, which contain enterochromaffin cell markers, likewise express TRPA1.

65TRPA1 was originally identified as a cold-activated channel, a role that later provoked some controversy as TRPM8 is attributed a greater role in this respect than TRPA1. Most remarkable, however, is that TRPA1 turned out to be a multimodal sensor not only for various spices but also for an extensive list of endogenous and exogenous irritant chemicals including a number of drugs in clinical use (Table 1). Many of these pungent and irritating chemicals are reactive electrophiles which need to be recognized and avoided because of their potential to damage tissues through modification of nucleic acids, proteins and other biomolecules.

66Its sensory modalities (Table 1) place TRPA1 specifically in a position (1) to taste spicy compounds present in mustard, horseradish, wasabi, galangal, black pepper, garlic, onion, cinnamon, ginger, oregano, wintergreen and clove (Table 1), (2) to detect toxic environmental stimuli (Table 1) such as ozone, tear gas, nicotine, formaldehyde, acrolein, iodoacetamide, methyl p-hydroxybenzoate (an antibacterial added to food, pharmaceutical and cosmetic products), irritating drugs (e.g., acetaminophen metabolites, fenamate nonsteroidal antiinflammatory drugs, isoflurane, propofol, dihydropyridines, and clotrimazole), and industrial chemicals (e.g., methyl isothiocyanate, and styrene), and (3) to survey the alimentary canal for potentially deleterious conditions arising from the presence of alkalosis, H2S, nitric oxide, and oxidative insult products such as 4-hydroxy-2-nonenal, H2O2, and acetaldehyde (Table 1).

67Although there is emerging evidence that TRPA1 contributes to mechanosensation in the gut, it appears that TRPA1 does not operate as primary mechanosensor. The slowly adapting currents which indentation of the soma membrane causes primarily in isolectin B4-negative DRG neuron cultures are suppressed by the TRPA1 blocker HC-030031 and are absent in TRPA1 knockout mice. Analysis of TRPA1-transfected HEK293 cells, however, shows that TRPA1 alone is not sufficient to confer mechanical sensitivity. It would appear, therefore, that TRPA1 functions as a secondary transducer of primary mechanosensors that await to be identified.

68In the oral cavity, TRPA1 mediates the taste of many spices (Table 1) and appears to contribute to the stinging and pungent sensations caused by carbon dioxide and carbonated beverages. Once it enters sensory nerve fibers, carbon dioxide causes intracellular acidification, which in turn activates TRPA1.

69Many studies have shown that activation of TRPA1 in the esophagus and GI tract causes excitation of primary afferent neurons in the vagus, splanchnic and pelvic nerves. TRPA1 agonists such as allyl isothiocyanate (AITC) excite vagal afferent C-fibers, but not Aδ-fibers, in the guinea-pig esophagus and spinal afferent neurons in the murine colon. Accordingly, intracolonic administration of TRPA1 agonists to rodents induces c-Fos expression in the spinal cord and elicits nociception as deduced from the occurrence of a visceromotor response and referred pain. Intraductal administration of a TRPA1 agonist to the murine pancreas likewise causes expression of c-Fos in the spinal cord. These nociceptive reactions to TRPA1 stimulation in the digestive system are inhibited by the TRPA1 blocker HC-030031, by intrathecal pretreatment with a TRPA1 antisense oligodeoxynucleotide or by TRPA1 gene knockout.

70TRPA1 is required for normal mechano- and chemosensory function in distinct subsets of vagal, splanchnic and pelvic afferent neurons supplying the murine GI tract. Specifically, spike discharges in response to punctate mechanical stimulation in splanchnic afferents with receptive fields in the mesentery and serosa and in pelvic afferents with receptive fields in the mucosa, muscularis and serosa are attenuated in TRPA1 knockout mice, while the responses of pelvic afferents to muscle stretch remain unabated. Despite these findings there is at present uncertainty as to whether TRPA1 contributes to normal mechanonociception in the colon, given that the visceromotor response to noxious colorectal distension has been reported to be reduced or to remain unchanged in TRPA1 knockout mice. In contrast, there is pharmacological evidence that in the rat stomach TRPA1 plays a role in mechanical nociception, given that the pseudoaffective pain reaction (contraction of the acromiotrapezius muscle) to gastric distension is attenuated by intrathecal administration of a TRPA1 antisense oligodeoxynucleotide as well as by intrathecal or intraperitoneal injection of HC-030031.

71TRPA1 is colocalized with TRPV1 in many afferent neurons, and there is evidence that the two TRP channels interact with each other in the sensory function of peripheral nerve fibers. For instance, the mechanical desensitization which capsaicin induces in splanchnic, but not pelvic, colonic afferents is absent in TRPA1 knockout mice, while the excitatory effect of capsaicin on these nerve fibers remains unaltered by TRPA1 gene deletion. The interaction between TRPV1 and TRPA1 is thought to be related to phosphatidyl 4,5-bisphosphate which is a factor relevant to the function of both TRP channels.

72In addition to stimulating afferent neurons, TRPA1 agonists sensitize colonic sensory neurons to mechanical stimulation and enhance the visceromotor response to colorectal distension, these effects being prevented by TRPA1 gene knockout or HC-030031. The ability of AITC to sensitize mechanosensitive afferent neurons may be related to its effect to enhance the level of TRPA1 in the colon. This argument is supported by a study in which AITC-induced irritation of the colon in newborn mouse pups results in a long-lasting hypersensitivity to colorectal distension as examined in the grown-up animals. The enhanced visceromotor response to colorectal distension seen in these experiments is thought to be driven by an early increase in colonic neurotrophin expression, which leads to a permanent increase in the percentage of TRPA1-positive DRG neurons, whereas the number of TRPV1-positive DRG cells remains unaltered.

73There is substantial evidence that TRPA1 contributes to the GI mechanical hypersensitivity that is associated with GI inflammation and stress. For instance, colitis induced by trinitrobenzene sulfonic acid (TNBS) in rats causes upregulation of TRPA1 in DRG neurons innervating the rat colon and enhances the visceromotor response to colorectal distension, this effect being reduced by intrathecal pretreatment with a TRPA1 antisense oligodeoxynucleotide. Likewise, the effect of TNBS-induced colitis to enhance spinal neuron activation and the visceromotor pain reaction to colorectal distension is absent in TRPA1 knockout mice. This observation is consistent with the ability of TNBS-induced colitis to increase the ability of TRPA1 agonists to sensitize mechanosensitive afferents in the splanchnic and pelvic nerves. Apart from GI inflammation, stress is also known to give rise to visceral hypersensitivity. The abdominal withdrawal reflex to colorectal distension, which is enhanced following chronic exposure to water avoidance stress for 10 days, is associated with a significant upregulation of TRPA1 in DRG neurons supplying the colon, but it has not yet been demonstrated whether TRPA1 is causally involved in stress-evoked visceral hyperalgesia.

74Under conditions of inflammation, TRPA1 appears to take over a significant role as secondary transducer of several proinflammatory mediators such as bradykinin and endogenous PAR-2 agonists. Pharmacological and gene knockout experiments indicate that TRPA1 mediates the bradykinin-induced mechanical sensitization of splanchnic afferents innervating the colonic serosa of the mouse and the bradykinin-evoked mechanical hypersensitivity of vagal afferent neurons in the guinea-pig esophagus. Importantly, the activation of vagal and splanchnic afferents by bradykinin itself is independent of TRPA1. Mast cell activation and activation of PAR-2 are further stimuli that cause mechanical hypersensitivity of vagal afferent neurons in the guinea-pig esophagus via a TRPA1-dependent mechanism. Similarly, the ability of a PAR-2 agonist to enhance the visceromotor response to colorectal distension is abrogated by TRPA1 gene deletion, while the effect of a PAR-2 agonist to stimulate colonic afferents in the splanchnic nerves per se is unaltered in TRPA1 knockout mice.

75The role of TRPA1 is not limited to mechanical hypersensitivity, as TRPA1 also participates in inflammation-evoked chemical hypersensitivity in the digestive system. Induction of mild colitis with DSS causes colonic hypersensitivity to AITC as revealed by increased expression of c-Fos in the murine spinal cord. This hyperresponsiveness to colonic TRPA1 stimulation is prevented by HC-030031 as well as by morphine, which suggests that the spinal c-Fos response to colonic TRPA1 stimulation reflects colonic chemonociception. Similarly, TNBS-induced colitis in the rat amplifies the visceromotor response to intracolonic AITC in a TRPA1 antisense oligodeoxynucleotide-sensitive manner. TNBS-evoked colitis is associated with an upregulation of TRPA1 in DRG neurons supplying the colon, an effect that may involve neurotrophic factors formed under conditions of irritation and inflammation.

76AITC has long been used to elicit and examine neurogenic inflammation, and there is evidence that activation of TRPA1 on visceral sensory neurons gives rise to both vasodilatation and inflammation. Introduction of excess AITC into the murine colon induces colitis which is associated with an upregulation of TRPA1 and other factors of the intestinal nervous and immune systems. Similarly, intraductal administration of an AITC agonist to the murine pancreas causes inflammation and activation of spinal neurons, these effects being markedly attenuated in TRPA1 knockout mice. The pancreatitis evoked by caerulein is likewise ameliorated by TRPA1 deletion, as is the activation of spinal neurons and the pain behavior that accompany pancreatitis. Neurogenic inflammation does not necessarily have a deleterious effect on the tissue, given that AITC has a protective effect against experimentally induced gastric lesions in the rat, an effect that appears to involve endogenous prostaglandins. The inhibitory effect which AITC and icilin have on the restitution of wounded monolayers of the rat gastric epithelial cell line RGM1 does not seem to be mediated by TRPA1.

77Stimulation of TRPA1 on the murine neuroendocrine cell line STC-1 causes cell activation and release of CCK, and activation of TRPA1 on rat enterochromaffin cells, on the rat endocrine cell line RIN14B and on the human pancreatic endocrine cell line QGP-1 leads to release of 5-HT. These effects of TRPA1 stimulation may be of functional relevance if the broad sensory spectrum of both TRPA1 and GI endocrine cells and the functional implications of CCK and 5-HT in digestion and GI signaling to the brain are considered. However, these implications of TRPA1 await to be explored and analyzed both in vitro and in vivo.

78Stimulation of TRPA1 in the gut has distinct effects on GI motor activity, the effects of TRPA1 stimulation being region- and species-dependent. Although intracolonic AITC (2.5%) does not alter compliance of the murine colon in vivo, isolated segments of the murine colon, but not small intestine, contract in response to AITC (30–100 μM) and allicin, the response being depressed by tetrodotoxin throughout the colon, whereas atropine inhibits the AITC-induced contraction in the distal colon only. In the guinea-pig ileum, AITC is able to evoke contraction via a mechanism that involves 5-HT and activation of 5-HT3 receptors, which suggests that the AITC-induced release of 5-HT from enterochromaffin cells has an impact on the control of intestinal motility.

79While the effect of TRPA1 stimulation in the intestine is to facilitate motor activity, gastric emptying in the rat in vivo is delayed by the TRPA1 agonists AITC and cinnamaldehyde through an action that also involves 5-HT and activation of 5-HT3 receptors. Administration of TRPA1 agonists to the dog in vivo stimulates motor activity in the gastric antrum and jejunum, and elicits giant migrating contractions in the colon. Activation of TRPA1 also has a prokinetic effect in the murine colon, in which AITC is able to antagonize atonic constipation induced by clonidine as well as spastic constipation induced by loperamide. However, the physiological implications of TRPA1 in the nerve circuits and smooth muscle mechanisms governing GI motility remain largely unexplored. Indirect evidence suggests that TRPA1 participates in the cold-induced contraction of the rat colon, but the physiological relevance of this finding is not yet possible to judge.

80A point mutation in the S4 transmembrane segment of the TRPA1 gene appears to be responsible for an autosomal-dominant familial episodic pain syndrome that is triggered by fasting and physical stress. Functional analysis shows that this mutation in the TRPA1 channel is associated with a 5-fold increase in inward current following activation and with enhanced secondary hyperalgesia caused by AITC. It awaits to be explored whether function-relevant mutations in the TRPA1 gene are relevant to GI disease.

81The properties of TRPA1 as a multimodal sensor for a huge variety of irritant chemicals (Table 1) indicate that TRPA1 blockers could be used to manage chemical irritation and pain. In addition, the emerging pathophysiological implications of TRPA1 suggest that both TRPA1 agonists and blockers hold therapeutic potential in particular disease entities. There is good reason to surmise that TRPA1 blockers could be of particular benefit in inflammation-induced hyperalgesia in the GI tract. TRPA1 blockers may also be of use in pancreatitis and inflammatory bowel disease. The discovery that activation of TRPA1 by some drugs contributes to their adverse effect profile (Table 1), which may include the alimentary canal, provides a rational basis for developing drugs that lack this unwanted property.

82TRPA1 blockers such as HC-030031 have already been characterized, and many other compounds are in the pipeline. The interaction between TRPA1 and TRPV1 in the excitation and/or sensitization of nociceptive nerve fibers hints at the possibility that combined TRPA1 plus TRPV1 blockers may have superior efficacy compared with blockers that target TRPA1 or TRPV1 alone. It awaits to be explored whether interference with TRPA1 disturbs temperature homeostasis, as it has been found with TRPV1 blockers. This limitation cannot be neglected, given that intragastric administration of the TRPA1 agonists AITC and cinnamaldehyde increases thermogenesis and reduces heat diffusion in mice.

83Apart from TRPA1 blockers, TRPA1 agonists also may have therapeutic potential. The large variety of TRPA1 agonists available to date (Table 1) can be roughly categorized in two classes: (1) agonists that stimulate TRPA1 via a transient and reversible interaction with the channel, and (2) agonists with an electrophilic carbon or sulfur that bind covalently to cysteine residues of the cytoplasmatic N-terminus of TRPA1. A specific application of TRPA1 agonists is suggested by their prokinetic effect in the murine colon, which—if translated to humans—could be utilized in the management of both atonic and spastic motor stasis of the gut. TRPA1 agonists that will be useful in therapy certainly require optimization in their pharmacological profile, since the utility of the currently available TRPA1 agonists is limited by their irritative effects and high reactivity.

84Expression of TRPM4 and, particularly, TRPM5, is limited to taste, epithelial and endocrine cells. TRPM4 transcripts have been identified in pancreatic α- and β-cell lines of rodent origin, and TRPM5 has been localized to pancreatic islets of the mouse in which it is expressed by the insulin-secreting β cells. TRPM5 abounds on the receptor cells of the lingual taste buds as well as on other chemosensory cells, notably on epithelial taste cells of the gut. Both TRPM4 and TRPM5 mRNA are found in the human digestive system. TRPM5 occurs in the stomach, small intestine and colon of mice and humans, the channel occurring in solitary brush (also termed tufted or caveolated) cells at the surface epithelium of the small and large intestine and in endocrine cells of the duodenal glands. Endocrine cells isolated from the rat stomach also stain positively for TRPM5.

85Functional evidence suggests that TRPM4 contributes to the membrane depolarization and increase in the intracellular calcium concentration that are relevant to glucose-, arginine vasopressin- and glibenclamide-evoked secretion of insulin from pancreatic β-cells and perhaps glucagon from pancreatic α-cells. TRPM5 is likewise thought to play a major role in the glucose-dependent release of insulin from pancreatic β-cells. Specifically, the effect of glucose to release insulin is significantly reduced in TRPM5 knockout mice, which impairs glucose tolerance.

86The Ca2+-gated TRPM5 channel on the receptor cells of the taste buds in the tongue has been established to contribute to signal transduction in the sweet, umami and bitter tastes. Bitter, sweet, and umami tastants bind to GPCRs which initiate a second-messenger signaling cascade involving activation of phospholipase Cβ2 and hydrolysis of phosphatidylinositol 4,5-bisphosphate to diacylglycerol and inositol trisphosphate. The subsequent increase in the intracellular Ca2+ concentration gates TRPM5 and causes depolarization of the taste cells and chemical transmission to gustatory afferent neurons. Real life gustatory experiences are often evoked by taste mixtures and determined by multiple interactions between different taste stimuli. One such interaction is the suppression of sweet taste transduction by bitter tastants such as quinine. Quinine and its stereoisomer quinidine have been found to block TRPM5 currents and to inhibit the nerve impulses of sucrose-stimulated gustatory nerves in mice. Since this quinidine-evoked inhibition of sweet taste signaling is absent in TRPM5 knockout mice, it would appear that TRPM5 plays an important role in bitter–sweet taste interactions.

87Apart from their occurrence in lingual taste cells, TRPM5 is widely expressed by other chemosensory cells, notably by the tufted epithelial cells of the GI tract. Given that the TRPM5-positive cells of the gut express most genes encoding proteins implicated in sweet, bitter, and umami taste signaling, it would appear that TRPM5 exert taste functions throughout the alimentary canal and contribute to the surveillance of the chemical composition in its lumen. This functional implication is corroborated by the finding that the expression of TRPM5 in the upper small intestine is inversely correlated with the blood glucose concentration in type 2 diabetes patients.

88Another task of TRPM5-positive taste cells in the digestive system may be to play local effector roles because they express markers of neuronal and inflammatory cells and can release biologically active messengers in response to stimulation. Solitary chemosensory cells in the upper small intestine of the murine coexpress TRPM5, β-endorphin, methionine enkephalin and uroguanylin. Stimulation of these cells with hyperosmolar solutions or glucose causes release of β-endorphin into the intestinal lumen but not blood serum, an effect that is blunted in TRPM5 knockout mice. In contrast, the intestinal release of β-endorphin evoked by acidic solutions or lipid exposure appears to be independent of TRPM5.

89Being a transducer of the sweet taste, TRPM5 has been suggested to have therapeutic potential in the management of obesity and related metabolic dysfunctions, given that taste signaling is a critical determinant of ingestive behavior. How efficacious such an approach may be is difficult to judge, however, because deletion of TRPM5 does not completely abolish the preference for sweets and energy-rich food. The reason for this finding is that there are TRPM5-independent transduction mechanisms for glucose tasting and sweet and calorie-rich nutrients can directly influence brain reward circuits that control food intake independently of functional taste transduction.

90As TRPM5 appears to be involved in glucose-dependent insulin secretion, TRPM5 dysfunction has been suggested to be a factor in the etiology of some forms of type 2 diabetes. If so, normalization of TRPM5 function may be envisaged as a target in diabetes management.

91TRPM6 and TRPM7 occur in GI epithelial cells and, as is true for TRPM7, in ICCs. In the murine gut, TRPM6 and TRPM7 abound in the brush-border membrane of the colon. Furthermore, TRPM7 mRNA and protein are expressed by AGS cells, the most common human gastric adenocarcinoma cell line. Apart from their association with epithelial cells, TRPM7 is present on ICCs of the myenteric plexus and circular muscle throughout the GI tract of mice and humans.

92Several data suggest that the formation of functional TRPM6 channels in epithelial cell membranes requires co-expression with TRPM7. There is indeed evidence for the formation of heteromeric ion channels consisting of both TRPM6 and TRPM7 subunits, but there is still controversy as to whether TRPM6 can also function on its own and TRPM7 is required only for trafficking of TRPM6 to the plasma membrane. TRPM6 and TRPM7 possess an atypical α-kinase domain and are involved in Mg2+ homeostasis, given that genetic and functional data indicate that both channel kinases play a critical role in the cellular Mg2+ absorption within the GI mucosa (especially in the colon) and in the reabsorption of Mg2+ in the kidney. Homozygous deletion of TRPM6 in mice is lethal, while heterozygous deletion of TRPM6 results in mild hypomagnesemia, a deficit that cannot be compensated for by a magnesium-rich diet.

93TRPM6 and TRPM7 channel function and colonic Mg2+ absorption are in fact controlled by many factors. For instance, the hypomagnesemia associated with dietary Mg2+ restriction in mice causes upregulation of TRPM6 expression both in the intestine and kidney. The expression of TRPM6, but not TRPM7, mRNA in the colon is likewise enhanced by a Mg2+-rich diet. Feeding of mice with long-chain inulin for 2 weeks, which increases Mg2+ absorption and enhances bone stores of Mg2+, results in a rise of TRPM6 and TRPM7 expression in the hindgut, while TRPM7 in the kidney is down-regulated.

94Mg2+ is critical for the growth and survival of certain cancer cells, notably the human gastric adenocarcinoma cell line AGS. Blockade of TRPM7 channels with La3+ or suppression of TRPM7 expression by siRNA inhibits the growth and survival of AGS cells.

95In addition to its epithelial function, TRPM7 plays a role in controlling intestinal motor activity. Electrophysiological, molecular biological, and immunohistochemical evidence indicates that TRPM7 participates in the nonselective cation current that underlies the pacemaker activity of ICCs in the GI tract. Thus, the electrophysiological and pharmacological properties of the nonselective cation current in ICCs of the mouse are identical to those of TRPM7, and the pacemaker activity of ICCs is inhibited by TRPM7-specific siRNA. TRPM7 expressed on ICCs is also likely to contribute to pacemaker activity in the human gut, since slow waves in the isolated muscle of the human small and large intestine are attenuated by blockade of TRPM7 channels with La3+.

96Non-functional mutations of TRPM6 are responsible for a primary disorder termed hypomagnesemia with secondary hypocalcemia. The expression of TRPM6 mRNA in the intestinal epithelium is reduced in patients suffering from cholera, a change that is thought to counteract the secretory response to the infection.

97The ability of TRPM7 channel blockade to inhibit the growth and survival of AGS cells suggests TRPM7 to be a potential target for the treatment of gastric cancer.

98TRPM8 is distributed to several cell types in the digestive system. Being menthol-sensitive, the TRPM8 subunit is expressed in the papillae of the tongue as well as by a distinct population of primary afferent neurons originating in the nodose-petrosal ganglion and DRG. Unlike that of TRPA1, the colocalization of TRPM8 with TRPV1 in sensory neurons is limited. Within the GI tract, TRPM8 has been localized to the muscle of the rat stomach and colon as well as to the liver. The expression of TRPM8 in the murine stomach and small intestine remains controversial, given that both positive and negative results have been reported.

99TRPM8 is activated by temperatures in the cool to cold range and makes an important contribution to cold-induced pain. Apart from low temperatures, TRPM8 is stimulated by various chemical entities including menthol, icilin, geraniol, L-carvone, isopulegol and linalool (Table 1). These sensory modalities of TRPM8 are likely to explain the refreshing taste of menthol. Also keeping with this property is the finding that intragastric administration of the TRPM8 agonist menthol induces thermogenesis in mice.

100Expressed by extrinsic primary afferent neurons and intrinsic systems within the GI tract, TRPM8 may play a chemosensory role in the alimentary canal. This contention is supported by the observation that the mixed TRPA1/TRPM8 agonist icilin is able to stimulate murine nodose ganglion neurons in culture. However, systematic studies of the chemosensory role of TRPM8-bearing primary afferent neurons in the gut have not yet been reported, and the precise functional implications of TRPM8 in the digestive system await to be explored. In the murine colon, menthol induces a long lasting relaxation of smooth muscle, which remains unaffected by tetrodotoxin. It is questionable if this motor effect of menthol is due to activation of TRPM8 because the presence of TRPM8 in the murine gut has not unequivocally been established. Whether the cooling-induced contraction of the rat gastric fundus and guinea-pig ileum involves TRPM8 has not conclusively been investigated.

101The polycystin TRP channel TRPP2 (also known as polycystic kidney disease-like ion channel, PKD2L1) and the related PKD1 polycystin L3 (PKD1L3) are expressed by a select class (type III) of taste chemoreceptor cells in the murine tongue, which are distinct from those mediating sweet, umami and bitter tastes. While TRPP2 has been reported to occur in the circumvallate, foliate and fungiform papillae of the tongue and in the palate, PKD1L3 has been localized to the circumvallate and foliate papillae only. PKD1, the gene that is mutated in ~85% of autosomal dominant polycystic kidney disease patients, has been described to occur in many tissues including the intestine.

102Although several members of the TRPV, TRPA and TRPC channel subfamilies are sensitive to alterations in extracellular pH, TRPP2 (PKD2L1) is the TRP channel involved in sour taste transduction. In order to form functional channels, TRPP2 needs to associate as heteromer with related proteins of the PKD1 polycystin family such as PKD1L3. The PKD1 polycystins are not included in the TRP channel family because they are large proteins with a very long N-terminal extracellular domain and 11 transmembrane domains that include a 6-transmembrane TRP-like channel domain at the C terminus.

103Both genetic and functional studies corroborate that TRPP2 plays an important role as sour taste receptor. In the taste buds of the murine tongue, TRPP2 is accumulated at the taste pore region where taste chemicals are detected. Coexpression of TRPP2 and PKD1L3 is necessary for the positioning of functional sour taste receptors at the cell surface, which is the case in the circumvallate and foliate papillae of the murine tongue. In native circumvallate taste cells expressing TRPP2 the pH threshold for acid-induced increases in the intracellular calcium concentration is around 5.0. Targeted ablation of TRPP2-expressing taste receptor cells results in a specific and total loss of sour taste transduction, whereas responses to sweet, umami, bitter and salty tastants remain unchanged. The relevance of PKD1, which occurs in the intestine, to GI function remains unexplored.

104The TRPP2-expressing sour taste cells also appear to contribute to the taste of carbonation. While the principal CO2 taste sensor is carbonic anhydrase-4, it remains to be determined whether TRPP2 plays a role in the downstream signaling of the carbonation taste.

105Distinct members of the classical or canonical TRP channels (TRPC) are widely distributed in the GI tract, the cellular sources being vagal afferent neurons, enteric neurons, smooth muscle cells, ICCs and epithelial cells. TRPC1, TRPC3, TRPC5 and TRPC6 have been localized to nodose ganglion neurons in which TRPC1 and TRPC6 are, at least in part, coexpressed with TRPV1 and TRPA1. The evidence for this colocalization rests on the finding that systemic pretreatment of rats with a neurotoxic dose of capsaicin causes a loss of TRPV1, TRPA1, TRPC1 and TRPC6 from vagal afferent neurons. In the mouse DRG, mRNA for all 7 types of TRPC subunits (TRPC1–TRPC7) has been found, with TRPC1, TRPC3 and TRPC6 being the most abundant.

106All 7 types of TRPC subunits have likewise been identified in the murine stomach. TRPC6 is expressed in human esophageal and gastric epithelial cells, while TRPC1 and TRPC5 have been localized to intestinal epithelial cells. In addition, TRPC4, TRPC6 and TRPC7 as well as splice variants of TRPC4 (TRPC4-α and TRPV-β) and TRPC7 are present in GI and vascular smooth muscle of the murine and canine alimentary canal, TRPC4 being the most abundant subunit. TRPC1, TRPC3, TRPC4, TRPC5 and TRPC6 mRNA have likewise been localized to circular muscle cells of the human esophageal body and lower esophageal sphincter, the levels of TRPC3 and TRPC4 mRNA being higher in the sphincter than in the body.

107The guinea-pig enteric nervous system expresses TRPC1, TRPC3, TRPC4 and TRPC6, which are differentially distributed to the myenteric and submucosal plexuses. While TRPC1 abounds in myenteric neurons with cholinergic and nitrergic phenotypes as well as in submucosal cholinergic and noncholinergic secretomotor neurons, TRPC3 is restricted to submucosal neurons containing neuropeptide Y. TRPC4 and TRPC6 are likewise preferentially expressed by noncholinergic secretomotor neurons of the submucosal plexus.

108The abundant distribution of TRPC channels to vagal afferent neurons, enteric neurons, smooth muscle cells, ICCs and epithelial cells in the gut points to multiple functional implications of this class of TRP channels. However, their role in the neural control of digestive functions has not yet been examined. There is, however, increasing evidence that TRPC channels are involved in signal transduction within the epithelium and smooth muscle of the GI tract. In addition, distinct TRPC channels play a physiological role in GI epithelial cell homeostasis and in the development of malignancies.

109The phenotype of TRPC1 knockout mice is characterized by a pronounced impairment of neurotransmitter-stimulated salivary gland fluid secretion, which suggests that TRPV1 is relevant to the secretory process in salivary gland acinar cells. In human colonic epithelial cells TRPC1 is involved in the signal transduction of the extracellular calcium-sensing receptor which is relevant to both intracellular calcium oscillation and inhibition of proliferation. Wounding of the mucosa causes translocation of stromal interaction molecule-1 to the plasma membrane, which in turn activates TRPC1-mediated calcium influx and promotes intestinal epithelial restitution. Induced TRPC1 expression facilitates, while inhibition of its expression by TRPC1-directed siRNA inhibits epithelial cell migration after wounding. In addition, induced TRPC1 expression sensitizes intestinal epithelial cells to apoptosis via calcium influx-mediated inhibition of nuclear factor-κB activity.

110TRPC1 and TRPC6 play a role in pancreatic and GI cancer development. In the pancreatic cancer cell line BxPc3, TRPC1 is a transducer of the action of tumor growth factor-β to stimulate cell motility, a process that is of relevance to tumor cell invasion. TRPC6, which can be activated both by receptor tyrosine kinases and GPCRs, appears to have a role in human esophageal and gastric cancer development. Treatment of the human gastric cancer cell lines AGS and MKN45 with the TRPC channel blocker SKF-96,365 arrests the cell cycle in the G2/M phase and suppresses the formation of gastric tumors in nude mice. Inhibition of TRPC6 likewise arrests the cell cycle of human esophageal squamous carcinoma cells in the G2 phase and inhibits tumor formation in nude mice injected with esophageal squamous carcinoma cells.

111In the exocrine pancreas, TRPC3-mediated calcium influx is attributed an adverse role in the acute inflammation induced by supramaximal stimulation or exposure to toxic bile acids and palmitoleic acid ethyl ester. Consequently, deletion of TRPC3 has been found to attenuate experimentally evoked acute pancreatitis.

112TRPC channels have been proposed to participate in intestinal pacemaker activity of ICCs which through electrical coupling drive slow wave activity in intestinal smooth muscle. The pacemaker current that initiates each slow wave derives from a Ca2+-inhibited, voltage-independent, nonselective cation channel in the ICCs. This channel exhibits properties similar to that reported for TRPC channels, notably TRPC4. However, slow waves in intestinal smooth muscle remain unchanged in TRPC4 knockout mice. Thus, there is no conclusive evidence that TRPC channels, notably TRPC4, play a physiological role in the pacemaker activity of ICCs.

113In contrast, distinct TRPC channels have been shown to operate as downstream transducers (effectors) of stimulated GPCRs such as muscarinic acetylcholine receptors. This mechanism has been suggested to apply to TRPC4, TRPC5 and TRPC6 which are activated by cholinergic stimulation of ICCs and smooth muscle in the murine stomach and intestine. Based on electrophysiological and pharmacological similarities, both TRPC4 and TRPC5 have originally been proposed to mediate the nonselective cation current evoked by carbachol in the murine stomach. A role of TRPC4 has been firmly established by the finding that the nonselective cation current evoked by muscarinic receptor stimulation is abolished by genetic deletion of TRPC4.

114Further studies using both TRPC4 and TRPC6 knockout mice indicate that TRPC4 underlies more than 80% of the muscarinic receptor-induced cation current in intestinal smooth muscle, while the residual current is mediated by TRPC6. Following TRPC4 knockout, the carbachol-induced membrane depolarization and the atropine-sensitive contraction elicited by acetylcholine release from excitatory motor neurons are inhibited, these effects being aggravated by additional deletion of TRPC6. In addition, intestinal transit is slowed down in mice lacking TRPC4 and TRPC6. From these studies it is evident that TRPC4 and TRPC6 couple muscarinic acetylcholine receptors to depolarization and contraction of intestinal smooth muscle. This is true for both M2 and M3 muscarinic receptors which use different second messenger mechanisms owing to their coupling to Gi/o and Gq/11 signaling pathways, respectively.

115An implication of TRPC6 in human GI motor control is exemplified by the observation that a single nucleotide polymorphism in the promoter region of the TRPC6 gene and a missense variant in exon 4 of the TRPC6 gene may contribute to infantile hypertrophic pyloric stenosis. TRPC6 appears to have a role in cancer development, given that there is a marked upregulation of TRPC6 expression in esophageal squamous cell carcinoma and in epithelial cells of human gastric cancers.

116Cholinergic contraction of smooth muscle is a key determinant of GI motility. Recognition that TRPC4 and, to a minor degree, TRPC6 are transducers of the depolarizing action of M2 and M3 muscarinic acetylcholine receptors in GI smooth muscle raises the possibility to attenuate cholinergic motor effects while sparing other cholinergically controlled tissue functions. This opportunity would be particularly useful if pathological hypermotility in the GI tract is shown to be due to exaggerated TRPC4 and TRPC6 transduction.

117Blockade of TRPC1 and TRPC6 also holds considerable potential in GI oncology, given that TRPC1 may promote pancreatic cancer cell invasion and TRPC6 seems to play a significant role in human esophageal and gastric cancer development.

118Distinct TRP channels have emerged as pleiotropic signal transducers in the digestive tract. They play physiological roles in taste, chemo- and mechanosensation, thermoregulation, pain and hyperalgesia, epithelial cell function and homeostasis, control of motility by neurons, ICCs and muscle cells, vascular function, and cancer development (Fig. 2). While the implications of some TRP channels, notably TRPA1, TRPC4, TRPM5, TRPM6, TRPM7, TRPV1, TRPV4, and TRPV6, have been elucidated to a considerable degree, the relevance of other TRP channels to digestive function awaits to be explored. The polymodal chemo- and mechanosensory function of TRPA1, TRPM5, TRPV1 and TRPV4 is of particular pathophysiological relevance, but there are several other aspects of GI function that are under the critical control of TRP channels, such as Ca2+ and Mg2+ absorption, transduction of muscarinic acetylcholine receptor activation to smooth muscle contraction, and cancer development (Fig. 2).

119TRP channels in the digestive system can operate (1) as molecular sensors (detectors or primary transducers) of chemical and physical stimuli, (2) as downstream or secondary transducers (or effectors) of cell activation induced by GPCRs, receptor tyrosine kinases or ion channels, and (3) as ion transport channels (Fig. 1). While some TRP channels operate only in one of these capacities, other TRP channels can act both as primary and secondary transducers. This dual role is seen with TRPA1, TRPV1 and TRPV4 which, on the one hand, are detectors of thermal, chemical and mechanical stimuli and, on the other hand, mediate cell responses to other stimuli, notably distension and proinflammatory mediators. A secondary transducer role is also typical of TRPM5 in its taste function and of TRPC4 in its role to transduce acetylcholine-evoked contraction.

120Pathophysiologically, TRP channels participate in many disturbances of GI function (Fig. 3). Changes in TRP channel expression and function are associated with GI inflammation, hyperalgesia and other disease identities, some of which are due to specific TRP channelopathies. These implications have raised enormous interest in the therapeutic exploitation of TRP channels as new drug targets in GI disease. As a result, extensive drug development programs have been initiated, especially with regard to TRPV1, TRPA1 and TRPV4 channel blockers. Fig. 4 summarizes the potential indications of TRP channel agonists and antagonists in the management of GI diseases and disorders. While the results of many preclinical tests involving TRP channel agonism, antagonism, knockdown or knockout nourish these expectations, clinical proof-of-concept data are not yet available, except for TRPV1 desensitization strategies. In the absence of conclusive information it is not possible to predict which TRP channels are the most promising drug targets. The translation from the preclinical validation to the clinical application of TRP channel-directed drugs is liable to meet many obstacles, given that TRP channels are widely distributed throughout the body and play many important physiological roles within and outside the GI system. The challenge, therefore, is to design TRP channel-directed drugs with a pharmacological profile that limits their action to the alimentary canal and specifically addresses the pathological functions of TRP channels while the physiological implications are spared.