How are transient lower esophageal sphincter relaxations triggered?
P.J. Hornby (New Orleans)
The tone of the lower esphageal sphincter (LES) and the contractions of the crural diaphragm  together provide a barrier to gastroesophageal reflux. Although it is apparent that low basal LES pressure alone, induced by atropine, is insufficient to provoke free reflux , failure of the gastroesophageal barrier results in reflux events. Since the LES relaxes during several physiological processes, it is important to understand the neural control of LES relaxation to gain insight into reflux disease. Swallowing involves LES relaxation, which can be accompanied by adaptive relaxation of the stomach to accommodate a bolus of food. In addition, transient LES relaxations (TLESRs) occur which do not directly follow a swallow and are independent of primary peristaltic activity of the esophageal body. Following a meal, the frequency of TLESRs is increased, and the resulting gastroesophageal reflux is generally cleared from the esophagus by primary or secondary peristalsis. In reflux disease there is an increase in frequency of TLESRs after a meal, which may be related to a greater acid reflux. Thus, an alteration in the triggering of TLESRs is now accepted as one of the key features in the development of gastroesophageal reflux disease [3, 4].
This review summarizes the current knowledge of how brain neurocircuitry controls swallowing and LES relaxations. It is hoped that by knowing the basic neurotransmitters and pathways governing LES relaxation, one can gain insight into the underlying factors that may be involved in triggering TLESRs. With this basic science insight then the ultimate task is to define which central receptor-mediated events may be modified to try to reduce the frequency of TLESRs and reverse or prevent reflux disease.
Vagal control of the LES
The precise stimuli and mechanisms underlying TLESRs are still controversial, although it is currently thought that TLESRs and experimentally-induced LES relaxations are controlled by "long-loop" neural feedback involving the vagus nerve. For example, TLESRs are triggered by sub-threshold pharyngeal stimulation  and LES relaxations are evoked by superior laryngeal  and vagus nerve  stimulation. In addition, distention of the stomach triggers TLESRs [8, 9] that are blocked by vagal cooling  and TLESRs are evoked via mechanical changes in the stomach , which are themselves controlled by a vagal afferent-vagal efferent pathways [11, 12]. Finally, the vast majority of sensory neurons innervating the LES occur in the vagal nodose ganglion . Two cautions are necessary before concluding that TLESRs are mediated by vagal pathways controlling LES relaxation. The first is that although these responses both involve the vagus nerve, it is not known whether the same vagal efferent fibers mediate the LES relaxation and TLESRs, although this is probably a reasonable prediction. Secondly, non-vagal pathways also contribute to control of the LES, since there is an afferent spinal innervation of the cat LES , and LES relaxation can involve vago-spinal pathways in ferrets . However, the consensus of literature is consistent with the idea that control of LES is primarily a vagal afferent-vagal efferent (vago-vagal) reflex.
Therefore, to gain insight into the mechanisms controlling TLESRs, it is first necessary to review the vagal neurocircuitry governing LES relaxation. Recently, the interest of the research community has focussed on the central nervous system for the study of extrinsic control of gastrointestinal function. This focus is reflected by the many symposia, books and reviews that address this topic. In a recent popular press article  the interaction between the brain and gut was dubbed the "Brain-Gut Highway" and described as a 2-way street that is necessary for normal reflexes, such as gastrointestinal transit, gastric adaptive relaxation, and esophagogastric relaxation.
The center of the integration of vagal control of the gastrointestinal tract is the dorsal vagal complex which is located in the dorsomedial hindbrain medulla (Figure 1). The term dorsal vagal complex comprises two major nuclei, one is the nucleus tractus solitarius (NTS) that receives the sensory input from the viscera, the other is the dorsal motor nucleus of the vagus (DMV) which contains the preganglionic motor output to the viscera. The dorsal vagal complex is actually a complex of many subnuclei arranged in columns extending from the spinal cord to the level of the fourth ventricle. An interesting aspect of this group of highly interconnected neurons is that it is relatively well-organized in terms of the viscerotopic representation of the gastrointestinal tract within regions of this structure. This means that sensory information from different parts of the gastrointestinal tract (including the esophagus - see below) enter into quite discrete areas of the NTS . Reflex changes in gastrointestinal function are then mediated by vagal motor neurons that reside in the DMV. The target of the vast majority of these neurons is the LES [17, 18] and stomach [19-21], although some preganglionic vagal neurons do innervate the duodenum [19, 22], cecum and proximal colon [23-25] and, to a lesser extent, the jejunum and ileum . The conclusion from these data is that the DMV innervates the entire gastrointestinal tract and can therefore exert direct control on all regions of it; however, the upper gastrointestinal tract is probably the most important target for vagal control based on the relative numbers of neurons innervating it.
Figure 1. Summary diagram of the dorsomedial medulla oblongata indicating the key nuclei in governing LES pressure. Open arrows indicate circulating agents, closed arrows indicate neuronal pathways.
Abbreviations: AP: area postrema; cc: central canal; cen: subnucleus centralis (esophageal afferents terminate); DMV: dorsal motor nucleus of the vagus; GABA: gamma aminobutyric acid; gel: subnucleus gelatinosus (gastric afferents terminate); mNTS: medial nucleus tractus solitarius (afferents from gastrointestinal, cardiovascular and respiratory systems); NO: nitric oxide; TS: tractus solitarius.
It is not known whether the underlying mechanisms of LES relaxation during a swallow are the same as those resulting in a TLESR. There are similarities indicative of some common underlying neural mechanisms, and it has been suggested that a TLESR is an incomplete swallow resulting from inadequate pharyngeal stimulation . It is therefore helpful to review what is known about the neurocircuitry controlling swallowing (Figure 2). In the NTS, sensory afferents from the esophagus terminate in a specific region, termed the subnucleus centralis . The cell bodies within this subnucleus project directly to somatic motor neurons within the nucleus ambiguus which govern the bucco-pharyngeal pattern generator for swallowing [27, 28]. Thus, there is a straightforward relay between information coming into the NTS from the peripheral sensory endings in the esophagus, to the motor neurons in the nucleus ambiguus that in turn innervate the muscles involved in deglutition (Figure 2).
Less is known about the neurocircuitry controlling LES relaxation; however, the data suggest that these neurons are also highly organized. In the early 1980's Barone et al.  stimulated the dorsal vagal complex by microinjection of L-glutamate, an excitatory amino acid, and produced diverse responses including decreased, increased or biphasic LES pressure responses . Later it was determined that vagal motor neurons innervating the LES are clustered within two distinct populations in the DMV, one rostral and one caudal to the obex [17, 31]. To be comparable with other species, obex is identified at approximately the calamus scriptorius , which is immediately behind the caudal edge of the area postrema. Based on the distance from the obex, the dorsal vagal complex is broadly divided into the caudal (caudal to obex/calamus scriptorius), intermediate (area postrema level) and rostral regions . When L-glutamate microinjections were made systematically into the DMV, increases in LES pressure were evoked from the rostral group and decreases in LES pressure were evoked from the caudal group  (Figure 1). The changes in LES and intragastric pressures in this study were not associated with changes in blood pressure whereas, in the study by Barone et al. , the evoked changes in LES pressure were associated with cardiovascular changes. It is therefore likely that stimulation of the NTS probably accounted for some of the changes since activation of this nucleus can result in both gastrointestinal and cardiovascular pressure changes .
In conclusion, it is likely that both peripheral ascending and central descending inputs related to esophageal function are integrated in the NTS, which then leads to activation of separate populations of LES excitatory and inhibitory preganglionic neurons in the DMV. Together the contribution of vagal inhibitory and excitatory neurons determine both the basal LES pressure, as well as the frequency amplitude and duration of spontaneous and swallow-induced LES relaxations. Therefore, the next question to address is the role of neurotransmitters in the dorsal vagal complex as well as the peripheral vagal terminations at the esophagus that may mediate LES relaxation.
Figure 2. Summary diagram of the key nuclei involved in swallowing with a focus on the components controlling somatic motor function.
Abbreviations are as listed in Figure 1. 5-HT: serotonin; ACh: acetylcholine;
a1: alpha 1 adreanoreceptors; EAA: excitatory amino acids
Neurotransmitters involved in vagal control of the LES
This section is divided up into two parts. The first part deals with neurotransmitters in the NTS that mediate afferent integration of swallowing and LES function and focuses on the effects of agents in the subnucleus centralis and medial subnuclei. The second part looks at the neurotransmitters that control vagal efferent outflow from the DMV to the LES.
There are very few experiments on the effects of neuroactive agents in the NTS on deglutition. This is unfortunate since these experiments are important to reveal the central integration of esophageal function at the level of the primary afferent input from the esophagus. Swallowing is induced by microinjection into the subnucleus centralis of
L-glutamate, or agonists of the N-methyl D-aspartate receptor subtype of excitatory amino acids. The esophageal component of the evoked swallow is blocked by N-methyl D-aspartate receptor subtype antagonists [35, 36]. The significance of this may be related to the fact that L-glutamate (or a closely-related substance) is a primary afferent neurotransmitter released in the NTS from esophageal afferents. Preliminary data suggest that L-glutamate may be a neurotransmitter in gastric primary afferents , and it is known to be within cardiovascular primary afferents entering the NTS from the carotid baroreceptors [38-40]. Therefore, excitatory amino acid release in the NTS in response to gastric distention or pharyngeal stimulation may evoke swallowing (Figure 2). Studies also show that microinjection into the NTS of acetylcholine , serotonin (5-HT)2 receptor agonists  and alpha1-adrenergic agonists  all stimulate pharyngeal and/or esophageal components of peristalsis (Figure 2). In addition, within the subnucleus centralis, the presence of somatostatin  and nitric oxide (NO) synthase [43, 44] within neurons projecting to the laryngeal and pharyngeal motor neurons of the nucleus ambiguus is well-described (Figure 2). Therefore, the relay between esophageal sensory information and the activation of swallowing may be nitrergic  and somatostatinergic . However, the extent to which this relay plays a role in the triggering of LES relaxation and swallowing remains to be determined.
Turning now to direct neurotransmitter control of the vagal preganglionic neurons, blockade of l-aminobutyric acid (GABA) tone, by microinjection of a GABAa antagonist into the DMV, decreases LES pressure in 71 % of DMV sites tested in cats . This suggests that there is a GABA drive on inhibitory vagal motor neurons that helps to maintain LES pressure (Figure 1). Therefore, one can speculate that a GABA interneuron relays tonic inhibitory information from the NTS to vagal motor neurons that inhibit LES pressure. Consistent with this scenario are preliminary data showing that DMV motor neurons are under tonic inhibitory control from the NTS  and that inhibitory synaptic currents in DMV neurons evoked by NTS stimulation are abolished by a GABAa antagonist .
Within the DMV the majority of preganglionic neurons contain acetylcholine, but there are also catecholaminergic  and nitrergic [49, 50] abdominal vagal preganglionics. The catecholamines are colocalized with acetylcholine in dorsal motor nucleus of the vagus neurons  and the functional significance of this in terms of LES pressure is unknown. Even the evidence that these catecholaminergic neurons project to the abdominal viscera has been disputed on technical grounds , and it is therefore unclear whether catecholamines in vagal preganglionic neurons play any role in control of the LES.
In contrast to the case for catecholaminergic neurons, nitrergic neurons form a separate population from the cholinergic neurons within the DMV [53, 54] and do probably control LES function. It is well known that NO released peripherally mediates LES [55, 56] and gastric [57-63] relaxation. Traditionally, these non adrenergic-non cholinergic (NANC) inhibitory pathways are thought to involve cholinergic preganglionic nerves that innervate nitrergic NANC myenteric (postganglionic) neurons in order to evoke muscle relaxation (Figure 3). However, the presence of NO synthase in neurons in the DMV [50, 64] leads to the question of whether this substance is present in vagal preganglionic neurons. We determined that NO synthase activity in the DMV was present in preganglionic neurons in the DMV innervating the gastrointestinal tract in rats by intraperitoneal injection of a retrograde tracer, fluorogold , and immunocytochemical techniques with fluorogold antiserum, combined with histochemical staining for NO synthase activity in the hindbrain. NO synthase-containing preganglionic neurons were concentrated in two populations within the DMV, one caudal to the obex, one rostral to the obex, with very few neurons double-labeled in the intermediate DMV between these two regions . This finding is interesting because it lends itself to speculation that NO may be a NANC neurotransmitter in vagal preganglionic neurons. Therefore, the presence of vagal nitrergic preganglionic neurons is a novel pathway that appears to exist in parallel to the more traditional view of vagal NANC inhibition.
Since L-glutamate stimulation of the caudal DMV results in LES relaxation in cats  it is possible that the NO synthase-containing preganglionic neurons in the caudal DMV are important for mediating LES relaxation. There is some experimental support for the contention that vagal inhibitory preganglionic motor neurons mediate LES relaxation. Inhibitory input to the esophageal smooth muscle is only attenuated by ganglionic blockade with hexamethonium , suggesting that preganglionic vagal cholinergic pathways innervating inhibitory NANC myenteric neurons cannot completely account for LES relaxation. However, it is still not clear whether these nitrergic preganglionic neurons target nitrergic NANC myenteric neurons to mediate smooth muscle relaxation. If this turns out to be the case, then vagal nitrergic preganglionic neurons are in a "command" position to control gastrointestinal relaxation (Figure 3).
Figure 3. Summary diagram of the functional antagonism of cholinergic excitation and nitrergic inhibition controlling gastric tone. Illustrated are two inhibitory NANC pathways, one involving preganglionic cholinergic neurons, the other involving preganglionic nitrergic neurons. It is not known whether both of these pathways innervate myenteric nitrergic neurons. A similar neurocircuitry may be proposed for control of LES pressure (see text for details). Abbreviations as shown in Figures 1 and 2, NANC: non adrenergic-non cholinergic; SP: substance P.
The primary target of vagal preganglionic neurons is the stomach and LES (as reviewed above). Therefore, our recent work has focussed on the neuroactive substances in the dorsal vagal complex that decrease intragastric pressure and may contribute to the mechanisms of gastric adaptive relaxation. While it is realized that the neurocircuitry may not be identical for control of intragastric pressure as for LES pressure, there is speculation that the proximal stomach and LES behave as a unit, since deglutition LES relaxation is combined with gastric adaptive relaxation to accommodate the bolus of food. In patients suffering from achalasia the gastric emptying of liquids is increased  and gastric adaptive relaxation is impaired . In addition, in up to 50% of patients suffering from gastroesophageal reflux disease there is delayed gastric emptying which may actually contribute to the increased gastric distention and the frequency of TLESRs in these patients . Thus, it is perhaps reasonable to extrapolate the results of these studies on central control of intragastric pressure to the effects of similar agents in the dorsal vagal complex to control of LES pressure.
If nitrergic preganglionic vagal motor neurons are in a "command" position to initiate LES and gastric relaxation, it should be possible to activate these neurons to evoke relaxation. Once this is accomplished, then the effects of various autonomic blocking agents on the evoked relaxation can be investigated. In general, in vitro techniques have been used to investigate the NANC transmitter(s) controlling gastric and LES inhibition [58, 59, 63, 69, 70] and it is believed that NO functionally antagonizes the excitatory effects of acetylcholine on the gastric smooth muscle . We have begun to investigate whether centrally-evoked gastric relaxation is dependent on NO, but did not want to stimulate the dorsal vagal complex directly since both cholinergic and nitrergic excitation would result in contractile and relaxant responses of the stomach, similar to the case for electrical field or vagal nerve stimulation. The mixed excitatory and inhibitory influences make it difficult to interpret the normal role of NO. For example, an antagonist of NO synthase in anesthetized rats not only abolishes the decrease in intragastric pressure evoked by vagal stimulation but also enhances the initial increase in intragastric pressure . Therefore, a mixture of autonomic antagonists is usually used in order to study the NANC component resulting in relaxation in vivo and in vitro. To overcome this problem and to attempt to stimulate a gastric relaxation that might mimic more closely the sequence of neurotransmitters leading to gastric relaxation than vagal nerve stimulation, we evoked a reduction of intragastric pressure by microinjection of substance P into the nucleus raphe obscurus (Figure 3). The nucleus raphe obscurus is a nearby hindbrain nucleus that has functional connections with the dorsal vagal complex [71, 72] and provides us with a reliable and repeatable model of vagally-mediated gastric relaxation . This decrease in intragastric pressure is observed without the addition of autonomic blockers, such as atropine and guanethidine, and allows us to study the pathways controlling gastric relaxation with a minimum of extraneous pharmacological intervention. The results of this study indicate that centrally-evoked gastric relaxation depends on simultaneous inhibition of cholinergic tone and activation of NO (and vasoactive intestinal polypeptide) pathways  (Figure 3). These studies underscore a role of vagal cholinergic tone and NO release in gastric relaxation, and similar studies on vagal cholinergic and NANC control of LES relaxation would be very helpful.
In addition to NO being an important NANC neurotransmitter controlling LES pressure, carbon monoxide may also have a role to play in LES function. This is based on the presence of the carbon monoxide producing enzymes haem oxygenase type 1 and 2 in the LES . Indeed, based on the presence of neurotransmitters or their receptors in the esophagus, many other substances may be involved in control of esophageal motor tone [76, 77]. However, assessment of the role of some of these agents leads to contradictory or species-dependent results. For example, cholecystokinin increases the frequency of TLESRs in dogs  but not in humans  and both excites and inhibits LES pressure in cats . In addition, different selective receptor subtype activation may also give different results regarding neurotransmitter control of LES function. For example, close intra-arterial infusion of 5-HT results in biphasic LES responses in ferrets . The 5-HT evoked relaxation was unaffected by vagotomy but was reduced by granisetron, indicating that
5-HT3 receptor mechanisms involve sympathetic pathways. However, in humans 5-HT3 antagonism had no discernible effects on LES function  although 5-HT4 antagonism by lintopride increases LES tone without affecting swallow evoked relaxation . Therefore, rather than list all the evidence for and against the role of various neurotransmitters in LES function, it is probably better to await more unified data and a clearer picture of these neurotransmitter mechanisms.
Conclusions and future perspectives
The fundamental question still remains - how are TLESRs triggered? Clearly feedback from from gastric distention  and pharyngeal stimulation  is processed in the NTS and is crucial in determining the presence of TLESRs. This vago-vagal visceral feedback could be in the form of direct (monosynaptic) afferent-efferent connections that synapse in the subnucleus gelatinosus of the NTS, similar to the situation for the stomach . Alternatively, the visceral feedback could be in the form of polysynaptic vago-vagal  and spino-vagal  synapses in the dorsal vagal complex. These latter multineuronal pathways allow more flexibility since the visceral information can be integrated with descending neuronal influences such as those from cortical, limbic and hypothalamic sites. However, the situation becomes more complex because, in addition to the neuronal pathways in the DVC, there are also non-neuronal feedback mechanisms from gastrointestinal tract in the form of circulating gastrointestinal peptides and hormones.
It is now evident that many gastrointestinal peptides that are released in response to food intake, such as pancreatic polypeptide and peptide YY, can act directly in the dorsal vagal complex to modulate gastrointestinal function [86-88]. It is beyond the scope of this paper to discuss the gastrointestinal effects of these peptides on visceral afferent pathways and the dorsal vagal complex, but a very important point needs to be made here. Circulating agents can influence gastrointestinal function due to a direct action on the dorsal vagal complex (Figure 1). This is because regions of the NTS are outside of the blood brain barrier and are thus exposed directly to circulating neuroactive agents. The evidence for this is that portions of the NTS have fenestrated blood capillaries and enlarged perivascular spaces  that have been shown to permit entry of large serum proteins  into the neuronal tissue. This is very similar to the situation for the nearby circumventricular organ, the area postrema, which mediates the emetic reflex in response to circulating noxious agents. This unique anatomy of the dorsal vagal complex makes it attractive as a potential pharmacological target for the manipulation of the reflex control of TLESRs because therapeutic agents circulating in the blood can have easy access to critical regions of the NTS. Gaining further knowledge about the neurotransmitters and neurocircuitry in the dorsal vagal complex controlling LES relaxation will enable us to take advantage of this opportunity.
17. Rossiter CD, Norman WP, Jain M, Hornby PJ, Benjamin S, Gillis RA. Control of lower esophageal sphincter pressure by two sites in the dorsal motor nucleus of the vagus. Am J Physiol 1990;259:G899-G906.
21. Norman WP, Pagani FD, Ormsbee HS III, Kasbekar DK, Gillis RA. Use of horseradish peroxidase to identify hindbrain sites which influence gastric motility in the cat. Gastroenterology 1985;88:701-705.
24. Altschuler SM, Firenci DA, Lynn RB, Miselis RR. Representation of the cecum in the lateral dorsal motor nucleus of the vagus and commissural subnucleus of the nucleus tractus solitarius in rat. J Comp Neurol 1991;304:261-274.
26. Altschuler SM, Bao X, Bieger D, Hopkins DA, Miselis RR. Viscerotopic representation of the upper alimentary tract in the rat: sensory ganglia and nuclei of the solitary and spinal trigeminal tracts. J Comp Neurol 1989;282:248-268.
28. Cunningham ET Jr, Sawchenko PE. A circumscribed projection from the nucleus of the solitary tract to the nucleus ambiguus in the rat: anatomical evidence for somatostatin-28-immunoreactive interneurons subserving reflex control of esophageal motility. J Neurosci 1989;9:1668-1682.
30. Ornsbee HS III, Barone FC, Lombardi DM, McCartney LC. Effects of hindbrain L-glutamate acid application on gastrointestinal motor function in the cat. In: Roman C, ed. Gastrointestinal motility: Proceedings of the ninth International symposium on gastrointestinal motility. Lancaster: MTP,1984:583-591.
31. Neil JP, Gonella J, Roman C. Localisation par la technique de marquage à la peroxydase des corps cellulaires des neurones ortho et parasympathiques innervant le sphincter sophagien inférieur du chat. J Physiol (Paris ) 1980;76:591-599.
37. Rogers RC, McCann MJ, Stephens RL. Evidence for glutamate as a neurotransmitter in the gastric mechanoreceptor afferents projecting to the nucleus of the solitary tract. Soc Neurosci Abstr 1990;16:865.
42. Menon MK, Kodama CK, Kling AS, Fitten J. An in vivo pharmacological method for the quantitative evaluation of the central effects of alpha-1 agonists and antagonists. Neuropharmacology 1986;25:503-508.
44. Gai WP, Messenger JP, Yu YH, Gieroba ZJ, Blessing WW. Nitric oxide-synthesizing neurons in the central subnucleus of the nucleus tractus solitarius provide a major innervation of the rostral nucleus ambiguus in the rabbit. J Comp Neurol 1995;357:348-361.
45. Washabau RJ, Fudge M, Price WJ, Barone FC. GABA receptors in the dorsal motor nucleus of the vagus influence feline lower esophageal sphincter and gastric function. Brain Res Bull 1995;38:587-594.
46. Zhang X, Fogel R, Renehan WE. Direct evidence that neurons in the dorsal motor nucleus of the vagus (DMNV) are inhibited by neurons in the nucleus of the solitary tract (NST). Soc Neurosci Abstr 1996;22:395.
50. Krowicki ZK, Sharkey KA, Cardoza SV, Nathan NA, Hornby PJ. Distribution of nitric oxide (NO) synthase in rat dorsal vagal complex and effects of microinjection of NO compounds upon gastric motor function. J Comp Neurol 1996;in press.
51. Armstrong DM, Gutman L, Haycock JW, Hersch LB. Co-localization of choline acetyltransferase and tyrosine hydroxylase within neurons of the dorsal motor nucleus of the vagus. J Chem Neuroanat 1990;3:133-140.
58. Allescher HD, Tougas G, Vergara P, Lu S, Daniel EE. Nitric oxide as a putative nonadrenergic noncholinergic inhibitory neurotransmitter in the canine pylorus in vivo. Am J Physiol 1992;262:G695-G702.
68. McCallum RW, Champion MC. Physiology, diagnosis and treatment of gastroesophageal reflux. In: McCallum RW, Champion MC, eds. Gastrointestinal motility disorders: diagnosis and treatment. Baltimore: Williams and Wilkins, 1990:135-162.
70. Lefebvre RA, De Vriese A, Smits JM. Influence of vasoactive intestinal polypeptide and NG-nitro-L-arginine methyl ester on cholinergic transmission in the rat gastric fundus. Eur J Pharmacol 1992;221:235-242.
77. Ny L, Alm P, Ekstrom P, Hannibal J, Larsson B, Andersson KE. Nitric oxide synthase-containing, peptide-containing, and acetylcholinesterase-positive nerves in the cat lower oesophagus. Histochem J 1994;26:721-733.
78. Boulant J, Fioramonti J, Dapoigny M, Bommelaer G, Buéno L. Cholecystokinin and nitric oxide in transient lower esophageal sphincter relaxation to gastric distention in dogs. Gastroenterology 1994;107:1059-1066.
79. Ledeboer M, Masclee AA, Batstra MR, Jansen JB, Lamers CB. Effect of cholecystokinin on lower oesophageal sphincter pressure and transient lower oesophageal sphincter relaxations in humans. Gut 1995;36:39-44.
85. Renehan WE, Zhang X, Beierwaltes WH, Fogel R. Neurons in the dorsal motor nucleus of the vagus may integrate vagal and spinal sensory information from the GI tract. Am J Physiol 1995;268:G780-G790.
87. Krowicki ZK, Hornby PJ. Hindbrain neuroactive substances controlling gastrointestinal function. In: Gaginella TS, ed. Regulatory mechanisms in gastrointestinal function. Boca Raton: CRC Press, 1995:277-319.