Neurotensin: a modulator of enteric cholinergic neurons in the guinea pig small intestine

Calcium regulated chloride permeabilities in primary cultures of rabbit colonocytes

To determine if calcium-dependent secretagogues directly act on epithelial cells to elicit Cl- secretion, their effects on Cl- transport and intracellular Ca(2+) concentrations ([Ca2+]i) were determined in primary cultures of rabbit distal colonic crypt cells. The Cl- sensitive fluorescent probe, 6-methoxyquinolyl acetoethyl ester, MQAE and the Ca(2+)-sensitive fluorescent probe, fura-2AM were used to assess Cl- transport and [Ca2+]i, respectively. Basal Cl- transport (0.274 +/- 0.09 mM/sec) was inhibited significantly by the Cl- channel blocker diphenylamine-2-carboxylate (DPC, 50 microM, 0.068 +/- 0.02 mM/sec; P < 0.001) and the Na+/K+/ 2Cl- cotransport inhibitor furosemide (1 microM, 0.137 +/- 0.04 mM/sec; P < 0.01). Ion substitution studies using different halides revealed the basal influx to be l- > F- > or = Cl- > Br-. DPC inhibited l- influx by approximately 50%, F- influx by 80%, Cl-influx by 85%, and Br- influx by 90%. Furosemide significantly inhibited influx of Br- (84%) and Cl- (81%) but not of F- and l-. The effects of agents known to alter biological response by increasing [Ca2+]i in other epithelial systems were used to stimulate Cl- transport. Cl- influx in mM/second was stimulated by 1 microM histamine (0.58 +/- 0.05), 10 microM neurotensin (2.07 +/- 0.32), 1 microM serotonin (1.63 +/- 0.28), and 0.1 microM of the Ca2+ ionophore A23187 (2.05 +/- 0.40). The Cl- permeability stimulated by neurotensin, serotonin, and A23187 was partially blocked by DPC or furosemide added alone or in combination. Histamine-induced Cl- influx was significantly inhibited by only furosemide. Indomethacin blocked histamine-stimulated Cl- permeability but had no effect on the actions of the other agents. These studies, focusing on isolated colonocytes without the contribution of submucosal elements, reveal that (1) histamine stimulates Cl- transport by activating the Na+/K+/2Cl- cotransporter via a cyclooxygenase-dependent pathway; (2) neurotensin, serotonin, and A23187 activate both Cl- channels and the cotransporter, and their actions are cyclooxygenase-independent.

Presynaptic muscarinic receptors modulate acetylcholine release from rat antral mucosal/submucosal nerves

The purpose of the present studies was to determine whether autoinhibition of acetylcholine release could be demonstrated in vitro from mucosal/submucosal neurons in rat antrum. Rat antral mucosal/submucosal tissues preloaded with [3H]choline were perifused and [3H]acetylcholine release measured under basal and stimulated conditions. Carbachol inhibited both spontaneous and evoked (electrical field stimulation, KCl) acetylcholine release from rat antral tissues: 1 x 10(-5) M carbachol inhibited basal [3H]ACh release maximally to -38.2 +/- 3.1% (P < 0.001 vs control). The nonselective muscarinic antagonist atropine enhanced both basal and stimulated acetylcholine release and abolished carbachol-induced inhibition of acetylcholine release. Pirenzepine, a muscarinic M1 receptor antagonist, inhibited acetylcholine release and did not alter carbachol-induced inhibition of acetylcholine release. In conclusion, acetylcholine release from rat antral mucosal/submucosal neurons is regulated negatively by a presynaptic feedback mechanism involving M2 and/or M3 receptors, while presynaptic M1 receptors facilitate release of neurotransmitter.

Effect of neurotensin on the canine gallbladder motility: in vivo and in vitro experiments

Neurotensin (NT) (10(-8)-10(-6)) exerted a dose-dependent increase in the tone and release of [3H]ACh in the guinea-pig gallbladder muscle strips but was inefficient in the canine gallbladder muscle strips. However, in conscious dogs NT (2.5-20 ng/kg intravenously (i.v.)) dose-dependently increased the gallbladder pressure. Similar was the effect of CCK8 (1-10 ng/kg i.v.) and carbachol (0.5-2 micrograms/kg i.v.). The NT- or CCK8-induced gallbladder pressure was inhibited by atropine (10-50 micrograms/kg i.v.) or hexamethonium (0.5-3 mg/kg i.v.). Somatostatin (1-2 micrograms/kg i.v.) or VIP (0.5-1 microgram/kg i.v.) also reduced or even abolished the NT- or CCK8-induced gallbladder pressure. The NT-induced increase of the tone of guinea-pig gallbladder preparations was accompanied by an increase of [3H]ACh release, suggesting the involvement of cholinergic innervation.

Functional and neurochemical evidence that neurotensin-induced release of acetylcholine from Auerbach's plexus of guinea-pig ileum is presynaptically controlled via alpha 2-adrenoceptors

The effect of neurotensin (NT) on [3H]acetylcholine release and contraction from isolated longitudinal muscle strip of guinea-pig ileum was examined. Neurotensin dose-dependently enhanced the release of [3H]-acetylcholine. This effect of neurotensin was inhibited by stimulation of alpha 2-adrenoceptors: noradrenaline, clonidine, xylazine or dexmedetomidine (alpha 2-adrenoceptor agonists) inhibited neurotensin-induced release of acetycholine (ACh) as well as the contractions, while CH-38083 or yohimbine (alpha 2-adrenoceptor antagonist) prevented this inhibitory effect. Our findings suggest that neurotensin may play a neuromodulatory role in the regulation of cholinergic neuronal activity in the gut and this modulatory effect is continuously controlled by the tonic activity of the sympathetic nervous system: endogenous noradrenaline release is capable of reducing the release of ACh and the consequent contraction of the gut enhanced by neurotensin.

Calcitonin gene-related peptide mediates capsaicin-induced neuroendocrine responses in rat antrum

BACKGROUND

Calcitonin gene-related peptide (CGRP) is a 37-amino acid peptide localized to primary sensory afferent nerves in the rat stomach. The actions of CGRP in regulating antral neuroendocrine function were examined in vitro through the use of capsaicin, an agent capable of evoking neuropeptide release from peripheral sensory nerve endings. These results were compared with the effects of exogenous CGRP and CGRP antagonist, CGRP8-37.

METHODS

Rat antral mucosal/submucosal fragments were incubated in either static or dynamic perifusion experiments. Media were assayed for gastrin, somatostatin, CGRP, and acetylcholine.

RESULTS

Capsaicin, like exogenous CGRP, stimulated antral somatostatin release and inhibited both gastrin release and acetylcholine discharge. Low dose capsaicin (1 x 10(-5) mol/L) caused significant stimulation of CGRP release: 33 +/- 0.2 vs. 14 +/- 1 pg/mL protein; P < 0.001. Tetrodotoxin blocked capsaicin-induced inhibition of acetylcholine release and prevented partially capsaicin-mediated stimulation of CGRP release. The CGRP receptor antagonist CGRP8-37 prevented capsaicin-induced D-cell stimulation and inhibition of G-cell secretion and cholinergic discharge.

CONCLUSIONS

The effects of capsaicin-induced changes in antral D- and G-cell secretion and acetylcholine discharge are due primarily to release of CGRP. Antral CGRP release from primary sensory afferent nerve terminals may act as a local effector substance to regulate antral neuroendocrine function.

Effect of neurotensin on contractile activity and [3H]acetylcholine release in cat terminal ileum during different postnatal periods

The effect of neurotensin (NT) on the contractile activity of circular and longitudinal strips from the terminal ileum of 15-, 30-, 60-day-old and adult cats as well as on the resting and electrically-evoked release of [3H]acetylcholine (ACh) was studied. Radioactivity was measured by liquid scintillation spectrometry and the effect of NT was evaluated by the S2/S1 ratio. In the circular muscle strips NT (1-100 nM) inhibited spontaneous contractions in all age groups. In the longitudinal strips the effect of NT was concentration- and age-dependent. NT at a concentration of 1 nM had no effect on the spontaneous activity in 15-day-old cats, but in the other age groups in 70-80% of the cats it inhibited spontaneous contractions. The response to 10 and 100 nM NT was either biphasic (relaxation followed by contraction) or inhibitory: in 15-day-old cats the response was biphasic only and with increasing age the percentage of strips responding with inhibition of the contractions increased. Neither substances affecting adrenergic and cholinergic transmission nor TTX changed the inhibitory response to NT. The contractile component of the biphasic response was TTX-resistant in all age groups and was significantly decreased by scopolamine in 60-day-old and adult cats. NT increased both resting and electrically-evoked release of [3H]ACh which was not changed by TTX. In the presence of the peptide the S2/S1 ratio increased as NT-induced [3H]ACh release in the strips of adult cats was higher than that in young cats.(ABSTRACT TRUNCATED AT 250 WORDS)

Validation of the antral mucosal/submucosal sleeve preparation: studies of gastrin and acetylcholine release in response to luminal stimulation

In the present study we developed an experimental model for direct assessment of antral endocrine cell and cholinergic neural responses to luminal stimulation. A sleeve of antral mucosal/submucosal tissue was prepared from rat antrum, mounted in perfusion chamber, and perfused in both luminal and submucosal compartments. Morphological and functional integrity of the antral sleeve were confirmed by histological examination and measurement of protein synthesis. Antral gastrin release was assessed in response to luminal stimulation with acid, peptone and distension. Luminal acid (pH3) inhibited basal gastrin release by -70.4% and luminal peptone stimulated gastrin release to 210% above control (p < 0.02). Distention of the antral sleeve by hydrostatic pressure (3-25cm H2O) caused stepwise and significant increase in gastrin release that was reversible. 3H-acetylcholine was stimulated significantly by KCl (56mM) to values twice control. In summary, these results establish the integrity and responsiveness of the antral sleeve to pharmacological and luminal stimulation. The antral sleeve may be a useful model in assessing antral function in response to luminal stimulation.

gamma-Aminobutyric acid localization and function as modulator of cholinergic neurotransmission in rat antral mucosal/submucosal fragments

gamma-Aminobutyric acid, a neurotransmitter in the central nervous system, has been shown to be present in and synthesized and secreted by rodent and feline myenteric plexus neurons. The aims of the present studies were to measure gamma-aminobutyric acid concentrations and synthesis and to establish cellular localization and uptake of gamma-aminobutyric acid by immunocytochemistry and autoradiography, respectively, within mucosal and submucosal tissues of the rat antrum. Direct demonstration of [3H]gamma-aminobutyric acid release and the effects of exogenous gamma-aminobutyric acid and muscimol, a GABA alpha agonist, on [3H]acetylcholine release from antral mucosal/submucosal fragments were examined in perifusion experiments. gamma-Aminobutyric acid content and synthesis, as reflected by glutamic acid decarboxylase activity, were present within antral mucosa at levels two to three times that of the body and muscular layers of both the gastric body and antrum. gamma-Aminobutyric acid was identified immunocytochemically, principally in mucosal epithelial cells of the antrum. Exogenous gamma-aminobutyric acid and muscimol were capable of stimulating acetylcholine release through a GABA alpha receptor-mediated mechanism that was abolished by tetrodotoxin. These results indicate that gamma-aminobutyric acid is present in and taken up by epithelial cells of the gastric antrum and that gamma-aminobutyric acid is capable of being synthesized by antral mucosal/submucosal tissues. Furthermore, these studies suggest that a peripheral gamma-aminobutyric acid mechanism that may modulate cholinergic neurotransmission and endocrine cell function exists within the antrum.

Direct and indirect stimulation of pancreatic exocrine secretion by neurotensin in anaesthetized dogs

1. The effects of neurotensin on pancreatic exocrine secretion were investigated both in the intact whole pancreas and in the isolated, blood-perfused pancreas ex vivo in anaesthetized dogs. 2. Intravenous (i.v.) injections of neurotensin (0.01-1 nmol/kg) elicited dose-dependent increases in the secretory rate of pancreatic juice without changes in plasma levels of cholecystokinin (CCK). The concentration of bicarbonate in the pancreatic juice induced by neurotensin was increased, but the protein concentration was scarcely changed. 3. The neurotensin-induced secretion was inhibited by SCH23390, a dopamine D-1 antagonist, but not by domperidone, phentolamine, propranolol, atropine, cimetidine, or L-364,718, a CCK antagonist. 4. Intra-arterial (i.a.) injections of neurotensin (0.1-3 nmol/kg) also elicited dose-dependent increases in the secretory rate of pancreatic juice flow, but did not change bicarbonate or protein concentration. The secretory activities were less effective and 1 nmol/kg of neurotensin i.a. was approximately equal to that of 0.03 nmol/kg of neurotensin i.v. 5. These results suggest that neurotensin mainly stimulates pancreatic secretion by acting indirectly. Neurotensin-induced secretion is, at least in part, mediated by endogenously released dopamine which activates dopamine D-1 receptors on the pancreas. In addition to its indirect action, neurotensin has a weak direct action to stimulate pancreatic secretion.

Neurotensin binding sites in porcine jejunum: biochemical characterization and intramural localization

Neurotensin is present in high concentrations in the mammalian gut, especially in enteroendocrine cells of the mucosa. Exogenous neurotensin has been shown to alter ion transport by the mucosa and contractile activity of intestinal smooth muscle. The purpose of this study was to determine the distribution of neurotensin binding sites within the intestinal wall. Initially, biochemical characteristics of [125I]neurotensin binding sites were determined within two preparations of the distal porcine jejunum: (1) the mucosa and submucosa, and (2) the circular and longitudinal muscle with their intramural plexuses. Ligand binding data for the preparation including the mucosa and submucosa indicated that [125I]neurotensin bound specifically to two sites having apparent equilibrium dissociation constants of approximately 0.046 and 0.37 nM. A binding site with a dissociation constant of approximately 0.38 nM was confirmed for the preparation of muscle and associated intramural plexuses. Xenopsin and neurotensin were equipotent to neurotensin in competing for these binding sites; neuromedin N was approximately 40 times less potent in the preparation of mucosa and submucosa. Receptor autoradiography was used to determine the distribution of [125I]neurotensin binding sites within the wall of the jejunum. Autoradiograms of [125I]neurotensin bound to cross sections of the proximal and distal jejunum showed that the highest densities of silver grains were associated with the internal submucosal ganglia, external submucosal plexus and myenteric ganglia. A moderate density of silver grains was associated with the circular muscle. The localization of neurotensin binding sites to submucosal ganglia is consistent with observations that neurotensin effects on active anion secretion by the mucosa are blocked by tetrodotoxin. Immunohistochemical localization of neurotensin in the porcine jejunum demonstrated a limited population of neurotensin immunoreactive cells within the mucosal epithelium. It is possible that neurotensin released from these cells in the mucosa as well as neurotensin-related peptides released from enteric neurons may be the endogenous ligands for the binding sites visualized in this study.

Novel autonomic neurotransmitters and intestinal function

A wide variety of substances, including amines and peptides, have been detected within the complex neuronal pathways of the enteric nervous system using immunohistochemical techniques. In this article we have discussed some of the more recent data on the effects of these substances on intestinal activity. We have also commented on the many difficulties associated with ascribing neurotransmitter status to individual compounds. The technique of immunoblockade of neurogenic functional responses has been used in an attempt to identify some of the putative neurotransmitter substances. The search for selective antagonists continues.

Effect of cholecystokinin octapeptide and somatostatin on the motility of guinea pig and canine gallbladder

Species differences have been observed in the effect of cholecystokinin octapeptide (CCK OP) on the canine and guinea pig gallbladder smooth muscle motility. 1. CCK OP was more potent stimulant in canine than in guinea pig gallbladder smooth muscles. Its pD2 values were 10 and 9.2, respectively. 2. The acetylcholine (10(-4) M)-induced maximum contractions in canine gallbladder muscle strips were by 50% lower as compared to the CCK OP (10(-8) M) maximum responses while in guinea pig gallbladder muscle strips the acetylcholine (ACh) maximum responses were by 20% lower than the CCK OP maximum responses. 3. CCK OP increased [3H]ACh release by 27% in canine gallbladder and by 40% in guinea pig gallbladder. 4. Somatostatin (SOM) had not any direct myogenic effect in guinea pig and canine gallbladder but it decreased [3H]ACh release from gallbladder intrinsic cholinergic neurons.

Effects of neurotensin-related peptides on the motility of the guinea pig oesophagus

Neurotensin, neuromedin N and xenopsin induced a monophasic and concentration-dependent contraction of the intact guinea pig oesophagus but kinetensin was without effect. The responses were completely abolished by tetrodotoxin and atropine but were unaffected by hexamethonium. The maximum response induced by neurotensin was reduced (20-30%) by somatostatin and by dynorphin-(1-13) in a naloxone-reversible manner. Neurotensin did not contract the isolated muscularis mucosae. The effects of neurotensin-related peptides on the motility of the oesophagus are mediated exclusively through the release of acetylcholine.

Effects of neurotensin on the motility of the isolated gallbladder, bile duct and ampulla in guinea-pigs

Neurotensin induced dose-dependent contraction in the isolated gallbladder, bile duct and ampulla of guinea-pigs, which were usually reduced by atropine and tetrodotoxin. In all cases, the neurotensin-induced contraction of the gallbladder was reversed to relaxation by indomethacin after administration of atropine and tetrodotoxin. The neurotensin-induced contraction of bile duct and ampulla was reduced by indomethacin, atropine and tetrodotoxin, and was slightly enhanced in some experiments after guanethidine administration. Ganglion-, alpha- and beta-adrenoceptor, serotonin- and histamine-blocking agents did not affect the neurotensin-induced contraction in any of the preparations. These results suggest the following; (1) the contractile effects are due to excitement of cholinergic neurons in the myenteric plexus of the biliary tract, (2) the direct action of neurotensin on the smooth muscle of the bile duct and ampulla results in a small contraction, (3) the contraction in the gallbladder is partly caused by stimulation of prostaglandin synthesis.

Calcium dependence of neurotensin stimulation of circular colonic muscle of the rabbit

The effect of neurotensin on smooth muscle contraction was compared in strips from rabbit proximal and distal circular colonic muscle. The effective dose for neurotensin stimulation that caused a 50% response in both tissues was similar (1.3 X 10(-10) M). The maximal isometric stress, however, was greater in the distal colon than in the proximal colon (p less than 0.01). Neurotensin stimulation of both proximal and distal colon was unaffected by tetrodotoxin, phentolamine, propranolol, naloxone, or atropine. Neurotensin-stimulated contraction was inhibited by "Ca2+-free" (pCa = 5.1) or La3+ buffer. Verapamil (10(-6) M) or nitroprusside (10(-4) M) decreased neurotensin stimulation of proximal and distal colon by approximately 40% (p less than 0.05). Removal of Ca2+ from the buffer inhibited stimulation of muscle contraction by high extracellular potassium [( K+]o) more than bethanechol stimulation (p less than 0.01). La3+ (1 mM) inhibited the contraction stimulated by bethanechol or increased [K+]o. Although verapamil inhibited contraction by bethanechol and increased [K+]o by approximately 50%, nitroprusside had no effect on the contraction mediated by these stimulants. 8-Bromo-guanosine 3',5'-cyclic monophosphate (cGMP) inhibited neurotensin, but not [K+]o or bethanechol-stimulated contraction. These data suggest (a) neurotensin stimulated colonic contractions at a concentration that is potentially physiologic, (b) neurotensin stimulated colonic smooth muscle directly without neural mediation, (c) neurotensin stimulation of colonic muscle is controlled by [Ca2+]o and [cGMP]i.

Mechanism of neurotensin depolarization of rabbit colonic smooth muscle

The aims of this study were (1) to measure the effect of neurotensin on the membrane potential of circular muscle of the distal colon of the rabbit and (2) to determine the mechanism by which neurotensin affects the membrane potential of this tissue. The membrane potential was measured with microelectrodes placed intracellularly and the double sucrose gap. Neurotensin (10(-11) M to 10(-7) M) dose-dependently decreased the membrane potential. The maximum decrease in membrane potential occurred with 10(-9) M neurotensin. The ED50 of neurotensin depolarization of the membrane potential was 0.87 +/- 0.33 X 10(-10) M. The frequency of the slow waves was unchanged after neurotensin. The voltage response to a constant current pulse decreased as the concentration of neurotensin increased. The amplitude of the voltage response after a 0.6 microA current pulse decreased by 6 +/- 0.5 mV after neurotensin (10(-7) M) compared to the Krebs control (P less than 0.05). Decreasing the [Na+]o to 0-23 mM did not affect the decrease in membrane potential after neurotensin. However, perfusion with a test solution containing no added Ca2+ or verapamil (10(-5) M) inhibited neurotensin depolarization of the tissue. Evidence was found that neurotensin depolarizes colonic circular smooth muscle, and the decrease in membrane potential is associated with an increase in conductance which is dependent on influx of Ca2+.

Release of gamma-aminobutyric acid and acetylcholine by neurotensin in guinea-pig ileum

The release of gamma-aminobutyric acid (GABA) and acetylcholine (ACh) from the strips of guinea-pig ileum was investigated in the presence of neurotensin. Neurotensin evoked the release of [3H]-GABA from the strips preloaded with [3H]-GABA, and the evoked release was Ca2+-dependent and tetrodotoxin-sensitive. Hexamethonium, scopolamine, [D-Pro2,D-Trp7,9] substance P and pretreatment with substance P did not alter the neurotensin-evoked release of [3H]-GABA. Pretreatment with neurotensin inhibited the release of [3H]-GABA evoked by neurotensin but not by high K+, thereby indicating that neurotensin induced a specific desensitization of its own receptor. These observations indicate that neurotensin may stimulate the GABAergic neurone through its own receptor. Neurotensin evoked the release of [3H]-ACh from strips preloaded with [3H]-choline and this release was Ca2+-dependent and tetrodotoxin-sensitive. The evoked release of [3H]-ACh was not affected by hexamethonium, scopolamine and [D-Pro2,D-Trp7,9] substance P. Bicuculline partly inhibited the neurotensin-evoked release of [3H]-ACh; thus neurotensin seems to induce a release of ACh partly through the release of endogenous GABA. All this evidence indicates that neurotensin induces release of GABA as well as ACh from the myenteric neurones of the guinea-pig ileum.

CSF neuropeptides in euthymic bipolar patients and controls

Levels of vasopressin, somatostatin, neurotensin, vasoactive intestinal peptide, corticotrophin-releasing factor and adrenocorticotrophin in CSF were determined in lithium-treated and unmedicated euthymic bipolar patients and controls, in a search for a trait marker in affective disorder. No group differences in levels of these peptides were found. Highly significant positive correlations were found among these peptides (with the exception of neurotensin), suggesting that their presence in CSF is functionally significant, as opposed to the result of random diffusion from the interstitial space of the brain.

Effect of neurotensin on amylase secretion from rat pancreas in vivo and in vitro

The effects of neurotensin, a tridecapeptide localized mainly in the gut, on pancreatic secretion are controversial. It is not clear whether the effects are mediated by direct or indirect effects. The present study was done to determine the effect of threshold doses of neurotensin in vivo in the rat as well as in vitro to clarify whether the action is direct or indirect. For in vivo studies the pancreatic duct of rats was cannulated and threshold doses of neurotensin and CCK-8 alone or in combination were infused as an intravenous bolus and pancreatic juice collected at 10-min intervals for 1 hr. For in vitro studies dissociated lobules and acini were incubated for 1 hr with 10(-11) -10(-6) M neurotensin with or without 10(-9) M CCK-8 (maximally effective dose). Amylase secretion in the pancreatic juice or in the medium was determined. Neurotensin and CCK-8 at a dose of 50 pM/kg increased pancreatic protein secretion in vivo. Neurotensin in combination with CCK-8 did not potentiate protein secretion. In the pancreatic lobules and acini, neurotensin 10(-11) -10(-6) M alone or in combination with CCK-8 (10(-9) M) did not stimulate amylase secretion above the basal level or that stimulated with CCK-8. The combined data indicate that the effect of neurotensin is not mediated by a direct action on the pancreatic acinar cells.

Neurotensin-induced colonic motor responses in dogs: a mediation by prostaglandins

The effects of systemic infusion of neurotensin (NT) were studied in four dogs fitted with strain-gauge transducers implanted on the ascending and descending colon. The motility index of the colon was enhanced for the duration of the 20 min NT infusion (20 pmol X kg-1 X min-1). Such stimulation was comparable to the colonic motor response elicited by feeding, i.e. the gastrocolic reflex. A period of hypomotility, lasting 40-90 min, occurred after either feeding or completion of NT infusion. Pretreatment with prostaglandin synthetase inhibitors, ketoprofene (KTP) and acetylsalicylic acid (ASA), reduced the magnitude of both the NT- and meal-induced hypermotility responses. The results suggest that the ability of NT to increase colonic motility may involve prostaglandin synthesis and that endogenous prostaglandin may exert a physiologic effect on colonic motor response to feeding. Moreover, these findings support a possible role for NT as one of the mediators involved in the non-neural mediation of the gastrocolic reflex.