The Shaggy Dog Story of Enteric Signaling: Serotonin, a Molecular Megillah

Abstract

Historically and quantitatively, the enteric site of serotonin (5-HT) storage has primacy over those of any other organ. 5-HT, by the name of “enteramine”, was first discovered in the bowel, and the gut produces most of the body’s 5-HT. Not only does the bowel secrete 5-HT prodigiously but it also expresses a kaleidoscopic abundance of 5-HT receptors. The larger of two enteric 5-HT stores is mucosal, biosynthetically dependent upon tryptophan hydroxylase1 (TPH1), and located in EC cells. Mechanical stimuli, nutrients, luminal bacteria, and neurotransmitters such as acetylcholine and norepinephrine are all able to stimulate EC cells. Paracrine actions of 5-HT allow the mucosa to signal to neurons to initiate peristaltic and secretory reflexes as well as to inflammatory cells to promote intestinal inflammation. Endocrine effects of 5-HT allow EC cells to influence distant organs, including bone, liver, and endocrine pancreas. The smaller enteric 5-HT store is biosynthetically dependent upon TPH2 and is located within a small subset of myenteric neurons. 5-HT is responsible for slow excitatory neurotransmission manifested primarily in type II/AH neurons. Importantly, neuronal 5-HT also promotes enteric nervous system (ENS) neurogenesis, both pre- and postnatally, through 5-HT_2B and especially 5-HT₄ receptors. The early birth of serotonergic neurons allows these cells to function as sculptors of the mature ENS. The inactivation of secreted 5-HT depends on transmembrane transport mediated by a serotonin transporter (SERT; SLC6A4). The importance of SERT in control of 5-HT’s function means that pharmacological inhibition of SERT, as well as gain- or loss-of-function mutations in SLC6A4, can exert profound effects on development and function of the ENS. Extra-enteric, TPH1-derived 5-HT from yolk sac and placenta promotes neurogenesis before enteric neurons synthesize 5-HT and contribute to ENS patterning. The impressive multi-functional nature of enteric 5-HT has made the precise identification of individual physiological roles difficult and sometimes controversial.

PubMed, the bibliographic service of the National Library of Medicine (USA), provides, as of August 2021, 6265 results for the query “serotonin and intestine.” That is a relatively large number of publications and indicates that scientific interest in that subject has been, if not extraordinary, at least substantial. It is thus surprising that a recent reviewer of an application to the National Institutes of Health (USA) for a research grant expressed incredulity that one might want to study a “brain chemical,” like serotonin (5-HT), in the gut. The popular association of 5-HT with “happiness”, or perhaps the strong focus of modern research on 5-HT in the brain, which has been the subject of ninefold as many papers as intestinal 5-HT, seems to have obscured, even for some of the scientists that study it, the many roles that 5-HT plays, not only in the gut but elsewhere outside of the CNS. In fact, enteric 5-HT is a polyfunctional molecule, acting as an endocrine hormone, a paracrine messenger, and a neurotransmitter [31]. It is also a mitogen [57] and growth factor that stimulates neurogenesis in the developing [28] and adult [48] enteric nervous system (ENS), as well as proliferation of the mucosal epithelium of the intestinal mucosa [37, 51]. The endocrine role of 5-HT links it to the control of osteocytes in bone deposition [43], the paracrine role to the initiation of peristaltic reflexes [14, 38] and inflammation [8, 23, 34, 50, 51], and the neurotransmitter role to the regulation of gastrointestinal (GI) motility [33, 41, 51]. To be sure, some of these putative functions of enteric 5-HT, such as its initiation of peristaltic reflexes, have been disputed [63], but it remains obvious that there was a good reason, or perhaps an abundance of reasons, for the biosynthesis of 5-HT in the GI tract to have evolved and thus to be studied. The enteric functions of 5-HT have been a complex tale, full of unexpected twists, turns, and reversals of fortune; thus it is very much a “shaggy dog story,” and the importance of 5-HT is what makes it a molecular megillah.

28.1. The History of 5-HT

The story of 5-HT in both the brain and the gut is not very old. It began in Italy during the great depression, well after the discoveries of the enteric nervous system (ENS) [2, 46, 53] and its ability to control to the behavior of the bowel independently of input from the brain or spinal cord [3–5, 67]. At that time, despite the depression and the rise of fascism in Italy, Vittorio Erspamer observed that material extracted from the wall of the bowel with acetone caused smooth muscle preparations to contract [25]. Erspamer further noted that the active principal reacted chemically as if it were an indole and that it was concentrated in the intestinal mucosa. Histochemical staining of gut sections, done to look for that indole, revealed that enterochromaffin cells of the intestinal epithelium contained an indoleamine that he named “enteramine” [11, 27, 69]. World War II intervened; however, Erspamer was distracted, and an entirely different trail of research would first reveal the actual structure of 5-HT.

The trail of research that would ultimately lead to the discovery of the structure of 5-HT had little or nothing to do either with the gut or the brain. It had to do with the quest of Irvine Page at the Cleveland Clinic to treat hypertension [73]. Page reasoned that there might be a vasoconstrictive substance in the blood of people whose blood pressure was abnormally high. A big problem for Page’s research was that blood clotted after it was drawn and that it had been known since 1846 that clotting released a vasoconstrictive substance into the serum. Page hired Maurice Rapport, a really good chemist who had just obtained a PhD degree, and Arda Green, a biochemist, who were given the task of getting rid of the interfering serum vasoconstrictor. Instead of doing that exactly, Rapport identified the molecule and proved the structure of 5-HT by synthesis [55]. The team of Rapport, Green, and Page chose the name “serotonin” for 5-HT because it was a serum vasoconstrictor. By 1953, Erspamer realized that “enteramine” and serotonin were the same thing [26, 27]; however, although one might argue that “enteramine” is a better name for a molecule that is most abundant in the bowel, it was too late. After Betty Twarog, working with Page, found that 5-HT was also present in the brain [68] and D. Wayne Woolley linked 5-HT to mental function [74], research aiming to determine the role of 5-HT in the brain-dominated science (as it still does) and the nomenclature had become permanent.

28.2. 5-HT and the Peristaltic Reflex

The study of the functional roles that 5-HT plays in the bowel began in earnest with a pioneering series of papers from the Oxford laboratory of Edith Bülbring [14–16, 19, 20]. She was interested in the initiation of the peristaltic reflex. This reflex, or something like it, was first observed by Bayliss and Starling who showed that pinching the bowel, or increasing pressure in the lumen of the intestine, elicited a reproducible series of movements of the gut consisting of contraction oral to the point of stimulation and relaxation distal to it, which they called the “law of the intestine” [3–5]. Because Bayliss and Starling were able to block that behavior with cocaine and nicotine but evoke it after they had severed all the extrinsic nerves running to the bowel, they attributed the activity to what they called the “local nervous mechanism” of the gut. Bayliss and Starling could do this because they were aware of prior discoveries made by Auerbach [2] and Meissner [53] that a very large intrinsic enteric nervous system (ENS) is present within the bowel. After the publication of the work of Bayliss and Starling, Paul Trendelenburg demonstrated intestinal behavior, which was identical to the “law of the intestine,” in preparations of gut that were isolated in vitro [67]. This observation of “the peristaltic reflex” in vitro put an exclamation point onto the conclusion of Bayliss and Starling that the behavior was due to the activity of the intrinsic ENS within the gut. Bülbring, who in her youth had lived in the same house as Trendelenburg, was anxious to determine how luminal stimulation is able to engage the ENS.

Bülbring and her associates devised a new procedure for demonstrating the peristaltic reflex. Instead of just recording the contraction of the longitudinal muscle or the pressure within the lumen of the gut, she measured both and, at the same time, the fluid the gut propelled [17]. This allowed Bülbring to conclude that she was truly looking at authentic peristaltic reflexes, which were propulsive, both in vitro and in vivo. The Bülbring group was able to use their improved preparation to follow up on the observations of Erspamer that showed that 5-HT is present in the intestinal mucosa, of Twarog and Wooley that 5-HT is present in neuronal tissue and neuroactive, and of Trendelenburg that increased intraluminal pressure evokes the peristaltic reflex. Bülbring first showed that intraluminal 5-HT enhanced peristaltic reflexes and lowered the threshold of pressure needed for their elicitation [14]. In contrast, serosal application of 5-HT inhibited or even blocked the peristaltic reflex [19]. In addition, she demonstrated that 5-HT is synthesized locally within the intestinal mucosa from 5-hydroxytryptophan and released from the mucosa when intraluminal pressure was raised and propulsive movement of the bowel was evoked [19]. These observations were important because Bülbring also showed that removal, asphyxiation, or anesthesia of the “mucous membrane” of the intestine impaired the activity that she identified as the peristaltic reflex, which was thus dependent for its initiation on the mucosa [20]. Bülbring further noted that primary afferent neurons are present in the bowel wall, she thought in the submucosal plexus, and she postulated that 5-HT is secreted from EC cells in response to mucosal stimulation and activates these intrinsic primary afferent neurons (IPANs) to initiate peristaltic reflexes [20]. Bülbring put these data together to suggest that increased intraluminal pressure releases 5-HT from EC cells, which stimulates the mucosal endings of IPANs that initiate peristaltic reflexes. The IPANs thus, in Bülbring’s hypothesis, act as the gateways to circuits within the ENS that control activity of the excitatory and inhibitory motor neurons that are ultimately responsible for the oral contraction and anal relaxation that comprise the “law of the intestine” or peristaltic reflex. Bülbring’s further studies confirmed that 5-HT is indeed released from EC cells in response to elevations of intraluminal pressure [16]. She subsequently went on to verify that responses of intact gut in vivo were identical to what she had demonstrated in vitro [15]. This work also established that very high concentrations of 5-HT could desensitize relevant 5-HT receptors and block its own action; however, Bülbring never was able to conclude, definitively in her mind, that mucosal 5-HT was essential for evoking peristaltic reflexes because reserpine, the agent that she employed to deplete 5-HT, was unable either to block 5-HT biosynthesis or reduce the 5-HT concentration to zero.

28.3. Did Bülbring Make a Mistake?

The deficiency that Bülbring recognized in her own work, that she could neither totally deplete the 5-HT stored in the bowel nor block its biosynthesis, has formed the basis of challenges to her hypotheses. No one, since Bülbring’s original work, has challenged the ability of 5-HT to evoke the peristaltic reflex, and skepticism has all been directed at the question of whether 5-HT signaling is “essential” for its manifestation. Boullin, for example [12], used a tryptophan-free diet to deplete 5-HT in rats up to a mean of about 90%. He found that the amplitudes of peristaltic reflexes in 5-HT-depleted animals were not significantly different from those in paired rats fed a normal diet. In fact, although no deficiencies in peristaltic activity were observed due to diet-induced 5-HT depletion, the effect of 5-HT on peristaltic reflexes was amplified in rats fed diets high in tryptophan, suggesting that an elevated level of endogenous 5-HT might add to the effect of exogenous 5-HT. Boullin nevertheless concluded that 5-HT was not essential for manifestation of the peristaltic reflex. Bülbring herself accepted Boullin’s conclusions and decided that mucosal 5-HT might be a modulator of the peristaltic reflex, acting to lower the pressure threshold needed to evoke the reflex although this role of 5-HT might not be an essential one [13].

More recent studies that have revisited the issue of the role of 5-HT in the peristaltic reflex have continued to focus on the question of whether 5-HT is “essential.” Studies have shown, for example, that a non-absorbed inhibitor of tryptophan hydroxylase (TPH), the rate-limiting enzyme in 5-HT biosynthesis, does not affect total GI transit time or colonic motility in mice [76]. In fact, even the genetic deletion of TPH1, which is the isoform of TPH found in EC cells, fails to alter total GI transit time, the time to expel a bead inserted into the rectum, gastric emptying, or small intestinal transit [47]. In contrast, deletion of TPH2, the isoform of TPH found in neurons, greatly slows parameters of intestinal motility, although it accelerates gastric emptying [47], perhaps by eliminating a neuronal role for 5-HT in vagal compliance (relaxation) of the stomach [18].

Methods utilized to study initiation of peristaltic reflexes have not been identical. There seems to be an underlying assumption that if one elicits the intestinal behavior that fits the description of the peristaltic reflex, then that is sufficient. Variations in stimuli have not been emphasized. It is likely, however, that more than one mechanism exists for engaging the microcircuits within the ENS that underlie peristaltic reflexes. Backup systems are common in biology; elimination of one pathway is often followed by activity in another that compensates for the absence of the pathway that has been deleted. In the case of the peristaltic reflex, there is evidence that, in contrast to what Bülbring reported, behavior of the gut that appears to be a peristaltic reflex can still be evoked, even after removal of the mucosa [44, 61, 62, 78]. This evidence, however, is also contradicted by evidence that Bülbring was right after all, and that removal of the mucosa does indeed abolish the peristaltic reflex [30, 39]. The apparent contradiction in observations has led groups to suggest that the opposing experimenters did not remove the mucosa properly, either damaging delicate nerve fibers underlying the mucosal epithelium, or incompletely removing the mucosa [60, 64]. There is no question, however, that no matter how adept one might be in removing the mucosa from the bowel, the preparation that remains after the mucosa has been removed is not a physiological one. A mucosa-free mammalian gut does not exist in nature. Whatever mechanism(s) the mucosa might have had to initiate peristaltic reflexes cannot operate after the mucosa has been removed; therefore, removal of the mucosa cannot provide definitive insight into how peristaltic reflexes are evoked when the mucosa and its connections to the ENS are intact.

28.4. 5-HT Is Essential for Mucosal Stimulation to Evoke the Peristaltic Reflex, But 5-HT Is Not Essential for Reflexes That Mechanosensitive Nerve Fibers Induce

Light brushing of the mucosal surface of an intact wild-type mouse gut evokes colonic migrating motor complexes (CMMCs), which are a form of peristaltic reflex; however, an identical stimulus fails to do so in mice that lack TPH1 (TPH1KO) [38]. Fecal pellets, moreover, are not propagated normally in the oral to anal direction in TPH1KO mice but may instead move irregularly or even reverse to move in the anal to oral direction. The colon of the TPH1KO mouse, furthermore, is elongated and dilated and contains grotesquely large fecal pellets. These many observations on the peristaltic reflex, taken together, suggest that more than a single mechanism exists that can engage the hardwired neuronal circuitry of the ENS that gives rise to the peristaltic reflex. In essence, at least two switches seem to exist that, when flipped, activate the machinery that drives peristaltic reflexes. One such switch or mechanism is mucosal stimulation, which operates indirectly through the secretion of 5-HT from EC cells. The secreted 5-HT excites mucosal projections of IPANs that serve as gateways to the circuitry of the ENS [38, 59]. The other mechanism is the initiation of peristaltic reflexes through the direct activation of mechanosensitive nerve fibers in the gut wall [63]. The mucosal pathway is 5-HT-dependent, while the direct activation of mechanosensitive nerve fibers is 5-HT-independent. Experimenters who stretch the bowel to evoke peristaltic reflexes thus find that the mucosa can be eliminated and that 5-HT is not essential. Experimenters who restrict their stimuli to the mucosa find that the mucosa cannot be eliminated and that 5-HT is essential. The gross normality of GI transit in TPH1KO mice probably reflects the ability of the bowel to compensate for the loss of the mucosal pathway to engage peristaltic reflexes; fecal pellets enlarge and stretch the gut, which eventually activates mechanosensors to trigger reflexes and propel luminal content. The abnormal elongated colon and massive fecal pellets of the TPH1KO mouse, no doubt, reflect the price these animals pay for not having the mucosal 5-HT-dependent pathway available.

28.5. Luminal Microbes Use EC Cells and 5-HT to Regulate the ENS

In vivo, the EC cells of the GI epithelium interact with luminal contents, including the massive microbiome of the bowel. Enteric microbiota, moreover, are not inert. Their presence enhances the ability of EC cells to synthesize and release 5-HT [56, 77]. As this happens, for example, when a germ-free mouse gut is colonized, the motility of the bowel is enhanced [24, 56, 77], an outcome that is consistent with the idea that EC cell 5-HT is linked to propulsive behavior of the gut. Strikingly, when the bowel of germ-free mice is colonized with a fecal transplant from conventionally raised mice, the neuroanatomy of the ENS changes in a way that suggests the occurrence of neurogenesis; moreover, GI transit also accelerates. These changes are associated with enhanced production of 5-HT, both in neurons and in EC cells of the intestinal mucosa [24]. The effects of depletion of endogenous 5-HT from EC cells and/or enteric neurons, as well as blockade of 5-HT₄ receptors, suggest that the action of the intestinal microbiota is to cause the release of 5-HT from EC cells and enteric neurons, which stimulates 5-HT₄ receptors that are neuroprotective and enhance maturation of the ENS, even in adult mice. The intestinal microbiome thus modifies the ENS and accelerates GI transit, not only through peristaltic reflexes, but also through trophic actions on enteric neurons [24].

28.6. 5-HT Is a Growth Factor for Enteric Neurons

The idea that 5-HT is a growth factor for enteric neurons predated the observations of the effects of enteric microbiota on EC cells and 5-HT. The earliest observations were made on enteric neurogenesis in vitro [28]. 5-HT was found to enhance neurogenesis when added to cultures of enteric neuronal precursors and promoted enteric neuronal development through an action of the 5-HT_2B receptor [28]. That observation was followed by evidence that the 5-HT₄ receptor might be at least as important as the 5-HT_2B receptor in mediating the ability of 5-HT to induce enteric neuroprotection and neurogenesis [24, 48]. The postnatal development of the ENS was found to be deficient in mice that lacked 5-HT₄ receptors, and 5-HT₄ agonists, like 5-HT_2B agonists, were powerful promoters of the development of enteric neurons from isolated neural crest-derived precursors [48]. Studies of the development of the ENS indicated that enteric serotonergic neurons occupy a strategic position in enteric neuronal development because they are among the first neurons to be born during ENS ontogeny [54]. Neurons that are born after the birth of serotonergic neurons undergo their terminal mitoses in the presence of serotonergic axon terminals [21]. Serotonin from the early-born neurons can thus contribute both to the ultimate numbers of these follower neurons and to the choices of their phenotypes. The physiological importance of 5-HT as a growth factor has been demonstrated by studies of the ENS in mice that lack TPH2, the isoform of TPH that is critical for 5-HT biosynthesis in neurons of the CNS [72] and ENS [47]. The deletion of TPH2, but not TPH1 (the TPH isoform in EC cells), leads to hypoplasia of the ENS and specific deficiencies of late-born neurons, such as those marked by their content of GABA, CGRP, and tyrosine hydroxylase. The effects of TPH2 deletion are essentially mimicked by TPH inhibition in neurons by para-chlorophenylalanine [24]. Interestingly, the effects of TPH2 depletion are not mimicked by administration of LP-920540 or LX1032, which are non-absorbed inhibitors of TPH that deplete mucosal, but not enteric neuronal 5-HT [52]. These observations indicate that enteric neuronal 5-HT is an important growth factor for the developing and mature ENS. As such, although GI transit is totally abnormal and slow in TPH2KO mice, it is impossible to know whether the defects in these animals are due to the absence of the neurotransmitter properties of 5-HT or to the absence of an appropriate number of neurons, or both. TPH2 also seems to be more important for ENS development than TPH1, although the studies of the effects of adding microbiota from conventionally raised mice to the bowel of germ-free animals [24] indicate that the mucosal as well as neuronal sources of 5-HT contribute to neuroprotection and ENS maintenance. It is possible that the effects of 5-HT from EC cells are not direct actions, but mediated indirectly through the ability, discussed above, of mucosal 5-HT to stimulate IPANs and thus engage the activity of ENS circuits.

28.7. SERT Plays a Critical Role in Enteric 5-HT Signaling

5-HT is released from the mucosa of the intestine in response to mechanical stimulation [6, 7]. Technological improvements have allowed mucosal 5-HT secretion to be followed for long periods of time in vivo and have confirmed that distension of the gut releases 5-HT in parallel with the peristaltic movements of the bowel [49]. Given the abundance of mucosal 5-HT and the observations of Bülbring that 5-HT blocks peristaltic reflexes and intestinal propulsion when it is applied to the serosal surface of the gut, it would seem to be necessary to compartmentalize mucosal 5-HT to its own layer of the gut and thus to protect the ENS from direct exposure to mucosal 5-HT. It is important to bear in mind that 5-HT is highly charged at a physiological pH; therefore, the molecule is not very membrane permeant in the absence of a specialized transporter. The enzymes that catabolize 5-HT, moreover, monoamine oxidase [11] and, in the bowel, glucuronyltransferase [32], are intracellular; thus unless 5-HT is transported into cells, it is not likely to be catabolized. In neurons, the serotonin reuptake transporter (SERT) catalyzes the uptake of 5-HT and is primarily responsible for the inactivation of 5-HT following its release at synapses and action on membrane receptors [10, 40]. The same transporter, SERT, is expressed in guinea pig brain and bowel [22] and rat gut [70], with a distribution of RNA that is consistent with its expression in enteric serotonergic neurons and mucosal epithelial cells. The enteric SERT of the guinea pig was cloned from the mucosal epithelium [22], and it is widely expressed throughout the guinea pig mucosal epithelium, although, in the rat, mucosal SERT is more heavily distributed in crypts [22]. The presence of SERT in the enteric mucosa enables the 5-HT that EC cells secrete to be inactivated rapidly, preventing excessive stimulation of mucosal receptors, their desensitization, as well as the overflow of 5-HT to swamp the ENS. In fact, the amplitude of EPSPs, recorded in cholinergic submucosal neurons (possible IPANs) following stimuli confined to the mucosa, is enhanced when SERT is inhibited with fluoxetine [22]. The number of submucosal neurons that respond to stroking of the mucosa, moreover, is also increased by the inhibition of SERT. Interestingly, the ability of mucosal stroking to excite submucosal neurons can be blocked with antagonists at 5-HT receptors, including 5-HT₃ and 5-HT₄ antagonists [29, 35, 42], as well as novel antagonists, such as anti-idiotypic antibodies that recognize 5-HT receptors [71] and a dipeptide of 5-HTP that is highly selective for actions of 5-HT on enteric neurons [65].

28.8. Inhibition of VMAT2

In assessing observations on mucosal or neuronal 5-HT signaling, it is important to consider the action of reserpine [58]. Reserpine is an old drug that, as discussed earlier, Bülbring first applied to the investigation of the peristaltic reflex [15]. Reserpine is an irreversible inhibitor of the vesicular monoamine transporter 2 (VMAT2) [9]. As such, it allows 5-HT biosynthesis to persist, although 5-HT (and other monoamine stores) is reduced to very low levels. Non-vesicular release of 5-HT, from the cytosol [45] possibly mediated by SERT acting in reverse, can still function, as Bülbring noted to her regret. To abolish the effects of endogenous 5-HT, therefore, it is necessary to delete TPH1 (for EC cells) or TPH2 (for neurons).

28.9. SERT Regulates Enteric Serotonergic Signaling

Because the inactivation of 5-HT is so SERT-dependent, alterations in SERT activity exert profound effects on the behavior of the gut and the development of the ENS [51]. The actions of SERT, moreover, are clinically relevant to autism spectrum disorder (ASD). GI disturbances are frequently seen in ASD [36]. Rare, hyperfunctional coding variants of SERT (encoded by SLC6A4) have been identified in ASD [51]. Expression in mice of the most common of these (SERT Ala56) increases 5-HT removal and leads to behaviors that resemble ASD. Mice that express SERT Ala56 also exhibit functional GI defects that are like those observed in TPH2KO animals, which lack neuronal 5-HT, including ENS hypoplasia, slow GI transit, diminished peristaltic reflex activity, and deficient proliferation of crypt epithelial cells. These observations suggest that the SERT overactivity is the functional equivalent of lack of 5-HT. In mice that lack SERT (SERTKO) and progeny of dams treated chronically with the SERT inhibitor, fluoxetine, an opposite phenotype occurs. Inability to inactivate 5-HT leads to its overactivity. These reciprocal phenotypes thus support the concept that serotonergic signaling is a powerful and critical regulator of ENS development. Disturbances of 5-HT signaling cause long-lasting abnormalities of GI function. The critical receptor for serotonergic signaling in ENS development appears to be 5-HT₄; thus, administration of prucalopride prevents the occurrence of SERT Ala56-associated GI perturbations. That observation implies that SERT overactivity deprives the 5-HT₄ receptor of neuronal precursors of their ligand, diminishing the trophic effects of 5-HT. Prucalopride can substitute for 5-HT, despite the SERT overactivity in SERT Ala56 mice, because prucalopride is not a substrate for SERT. It seems likely that GI and behavioral features of ASD are due to deficient 5-HT signaling during development. The safety and potentially adverse effects on the gut of SERT inhibition during pregnancy to combat depression therefore must be considered.

28.10. The Importance of Extraenteric TPH1 During Early Development

The developmental importance of 5-HT in the formation of the ENS and the role of microbiota [24] are certainly consistent with the idea that TPH1 is very important. The role of SERT in the compartmentalization of the mucosa [22], however, and the severe ENS defects seen in mice that lack TPH2 [47] cast a dubious light on the relative importance of the TPH1-derived 5-HT pool in EC cells. On the other hand, serotonergic neurons themselves appear to be sensitive to the trophic effects of 5-HT, which suggests that 5-HT might influence ENS development before the appearance of serotonergic neurons or EC cells. In the CNS, the placenta and especially the yolk sac provide an early source of 5-HT in the circulation that exerts profound effects on the patterning of the forebrain [75]. This 5-HT is TPH1-dependent, like that of EC cells. We have recently confirmed these data (Margolis and Gershon, personal observation) and found that the murine placenta and yolk sac are rich in TPH1 at E12/13, a time when little or no TPH2 is expressed in the brain or gut. As a result, TPH1 deletion is potentially able to disrupt ENS development and patterning, even if the effects are subtle and not as obvious as those due to deletion of TPH2.

28.11. Tryptamine

One tool that has been able to provide insight into the role of TPH1-derived 5-HT has been to make use of tryptamine to evoke the secretion of endogenous 5-HT. Tryptamine was demonstrated years ago to act on serotonergic neurons in a manner analogous to the action of tyramine and other indirectly acting sympathomimetic compounds on catecholaminergic neurons [66]. Tryptamine enters serotonergic axon terminals, and VMAT2 transports it into synaptic vesicles. Because tryptamine is a weak base, it collapses the pH gradient across vesicular membranes, which, in turn, causes 5-HT to flow out of vesicles into the cytosol. Tryptamine further activates trace amine-associated receptors (TAAR1)[1] and phosphorylates SERT, causing it to act in reverse and pump 5-HT into the extracellular space. Once released from terminals, 5-HT gains access to its receptors and reveals the effects of endogenous 5-HT release. These effects in the ENS include the mediation of slow transmission in type II/AH neurons [66]. Chronic exposure to tryptamine, furthermore, depletes endogenous 5-HT, thereby blocking the effects of nerve stimulation on type II/AH neurons, while permitting these cells to respond normally to exogenous 5-HT and to the endogenous release of acetylcholine. More recent studies have verified the action of tryptamine on the ENS and have shown that tryptamine is without effect, when it is applied to TPH2KO bowel, which lacks 5-HT. In contrast to the TPH2KO gut, TPH1KO bowel still responds to tryptamine but does so abnormally (Gershon and Margolis, personal observations). When indices of responses to tryptamine are compared in wild-type bowel to the TPH1KO gut, neuronal uptake of FM2–10, cytochrome oxidase activity, and Fos activation in wild-type bowel are all much greater in the wild-type bowel than in that of TPH1KO mice. Strikingly, activation of glial Fos in the TPH1KO gut is stronger after administration of tryptamine than in neurons. These observations are compatible with the supposition that the deletion of TPH1 leads to abnormal patterning of enteric neurites, even though the numbers of various enteric neurons may be close to normal.

28.12. Summary and Conclusions

• The gut bristles with 5-HT and 5-HT receptors.

  • The larger of two enteric 5-HT stores is TPH1-derived and is located in EC cells.

    • EC cells respond to mechanical stimuli, nutrients, luminal bacteria, and neurotransmitters, such as acetylcholine and norepinephrine.
    • Paracrine effects allow the mucosa to signal to neurons to initiate peristaltic and secretory reflexes as well as to inflammatory cells to promote intestinal inflammation.
    • Endocrine actions of 5-HT allow EC cells to exert effects on distant organs, including bone, liver, and endocrine pancreas.
  • The smaller 5-HT store is TPH2-derived and is located within a small subset of enteric neurons.

    • 5-HT is responsible for slow excitatory neurotransmission manifested primarily in type II/AH neurons.
    • 5-HT also promotes ENS neurogenesis, postnatally mainly via 5-HT₄ receptors; moreover, the effects of 5-HT on enteric neurogenesis are manifest during development and in adult life.

• Extra-enteric, TPH1-derived 5-HT from yolk sac and placenta promote neurogenesis before enteric neurons synthesize 5-HT and contribute to ENS patterning.