5-Hydroxytryptamine 4 Receptor Agonists Initiate the Peristaltic Reflex in Human, Rat, and Guinea Pig Intestine

Transcription

1 GASTROENTEROLOGY 1998;115: Hydroxytryptamine 4 Receptor Agonists Initiate the Peristaltic Reflex in Human, Rat, and Guinea Pig Intestine JOHN R. GRIDER, AMY E. FOXX ORENSTEIN, and JI GUANG JIN Departments of Physiology and Medicine, Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia Background & Aims: The peristaltic reflex induced by mucosal stimuli is mediated by intrinsic sensory calcitonin gene-related peptide (CGRP) neurons activated by 5-hydroxytryptamine (5-HT) released from enterochromaffin cells. The involvement of 5-HT 4 receptors was examined with selective 5-HT 4 agonists. Methods: Compartmented intestinal segments were used to measure neurotransmitter release and the mechanical components of the reflex. Results: In human jejunal and rat and guinea pig colonic segments, addition of the 5-HT 4 agonist HTF 919 elicited release of CGRP only into the compartment where the 5-HT 4 agonist was added; vasoactive intestinal peptide (VIP) was released only into the compartment where descending relaxation was measured, and substance P (SP) was released only into the compartment where ascending contraction was measured. The CGRP antagonist hc- GRP8-37 inhibited both mechanical responses by 75% 80%. Release of CGRP, VIP, and SP as well as ascending and descending responses were inhibited by selective 5-HT 4 but not by selective 5-HT 3 antagonists. Similar results were obtained with a different 5-HT 4 agonist, R However, HTF 919 was times more potent (median effective concentration, 10 nmol/l for peptide release and 5 nmol/l for mechanical responses) than R Conclusions: Selective 5-HT 4 agonists applied to the mucosa in nanomolar concentrations trigger the peristaltic reflex in human, rat, and guinea pig intestine. The peristaltic reflex can be elicited by muscle stretch and by mechanical or chemical stimulation of the mucosa. 1 4 The motor limbs of the reflex are identical regardless of the stimulus. The orad phase of the reflex consists of contraction of circular muscle and reciprocal relaxation of longitudinal muscle, whereas the caudad phase of the reflex consists of relaxation of circular muscle and reciprocal contraction of longitudinal muscle. 5,6 Relaxation of circular muscle is mediated by the inhibitory neurotransmitters, vasoactive intestinal peptide (VIP) 1,7 and pituitary adenylate cyclase activating peptide (PACAP), 8 and by nitric oxide generated in both nerve terminals and smooth muscle cells. 9,10 Contraction of circular muscle is mediated by the excitatory neurotransmitters, acetylcholine and the tachykinins, substance P (SP), and neurokinin A (NKA). 7,11 13 In contrast, the sensory limb of the reflex uses different pathways depending on the stimulus. The sensory pathway activated by mucosal stimuli is wholly intrinsic, whereas the sensory pathway activated by muscle stretch is mediated by extrinsic sensory neurons with cell bodies in the dorsal root ganglion. 1 The evidence of distinct pathways may be summarized as follows: (1) the pathways activated by mucosal stimuli and muscle stretch can be differentially desensitized 3,4,14,15 ; (2) simultaneous activation of both pathways results in potentiation 3,16 ; (3) removal of the mucosa eliminates the reflex elicited by mucosal stimuli but not the reflex elicited by circular muscle stretch ; and (4) surgical or chemical ablation of extrinsic sensory fibers eliminates only the reflex elicited by muscle stretch, 1 whereas 5-hydroxytryptamine (5-HT) antagonists eliminate only the reflex elicited by mucosal stimuli. 20,21 The claim that muscle stretch can induce a peristaltic reflex after chronic extrinsic denervation is based on the unlikely assumption that balloon distention from the serosal side does not induce concurrent mucosal stimulation. 22 The peristaltic reflex in these experiments was evaluated by measurement of the electrical correlates of ascending contraction and descending relaxation, i.e., ascending excitatory and descending inhibitory junction potentials. Both the extrinsic sensory neurons activated by muscle stretch and the intrinsic sensory neurons activated by Abbreviations used in this paper: GR A, (1-[2-(methylsulfonylamino)ethyl]-4-piperidinyl)methyl 1-methyl-1H-indole-3-carboxylate, maleate salt; hcgrp8-37, human calcitonin gene-related peptide 8-37; 5-HT, 5-hydroxytryptamine; LY , 1-methyl-N-(8- methyl-8-azabicyclo[3.2.1]-oct-3-yl)-1h-indazole-3-carboxamide maleate; NKA, neurokinin A; PACAP, pituitary adenylate cyclase activating peptide; R093877, 4-amino-5-chloro-2,3-dihydro-N-(1-[3- methoxypropyl]-4-piperidinyl)-7-benzofurancarboxamide monohydrochloride; SDZ , 2-methoxy-4-amino-5-chloro-benzoic acid 2-(diethylamino) ethyl ester; SDZ HTF 919, 5-methoxy-indole-3- carboxaldehyde amino(pentylamino) methylene hydrazone hydrogen maleate by the American Gastroenterological Association /98/$3.00

2 August HT 4 AGONISTS ELICIT THE PERISTALTIC REFLEX 371 mucosal stimuli contain calcitonin gene-related peptide (CGRP). 2,20,21 Release of CGRP from intrinsic sensory neurons is triggered by release of 5-HT from mucosal enterochromaffin cells. 20,21,23 In human intestine and rat colon, 5-HT acts on 5-HT 4 receptors located on CGRP neurons, whereas in guinea pig colon, 5-HT acts on both 5-HT 4 and 5-HT 3 receptors. Both 5-HT and CGRP are released in response to mucosal stimulation, and the release of CGRP is abolished by 5-HT 4 antagonists in human intestine and rat colon and by a combination of 5-HT 4 and 5-HT 3 antagonists in guinea pig colon. 20,21 In human intestine and rat colon, both 5-HT 4 and CGRP antagonists block the release of excitatory and inhibitory neurotransmitters and the motor components of the peristaltic reflex; in guinea pig colon, both 5-HT 3 and 5-HT 4 antagonists as well as CGRP antagonists block neurotransmitter release and mechanical response. The pattern suggests that 5-HT acts on 5-HT 4 receptors located on CGRP-containing sensory nerve terminals in all three species and also on 5-HT 3 receptors in guinea pig colon. In the present study, we used the selective 5-HT 4 agonists SDZ HTF ,25 (5-methoxy-indole-3-carboxaldehyde amino(pentylamino) methylene hydrazone hydrogen maleate) and R (4-amino-5-chloro-2,3-dihydro-N- (1-[3-methoxypropyl]-4-piperidinyl)-7-benzofurancarboxamide monohydrochloride) to further examine the role of 5-HT 4 receptors in initiating the peristaltic reflex in human small intestine and in rat and guinea pig colon. In all three species, both 5-HT 4 agonists caused a concentration-dependent release of CGRP at the site of stimulation as well as VIP release caudad and SP release orad to the site of stimulation. Release of VIP and SP was accompanied by circular muscle relaxation and contraction, respectively. Materials and Methods Compartmented Flat-Sheet Preparation of Guinea Pig and Rat Colon A three-compartment flat-sheet preparation previously described in detail 1,2 was used to measure neurotransmitter release and contraction or relaxation of circular muscle. In brief, 5-cm segments of mid-to-distal colon were opened and pinned mucosal side up to the base of a chamber and bathed in a Krebs-bicarbonate medium containing 118 mmol/l NaCl, 4.8 mmol/l KCl, 1.2 mmol/l KH 2 PO 4, 2.5 mmol/l CaCl 2, 1.2 mmol/l MgSO 4, 25 mmol/l NaH 2 CO 3, 11 mmol/l glucose, 0.1% bovine serum albumin, 10 µmol/l amastatin, and 1 µmol/l phosphoramidon. The sheets were separated into compartments by vertical partitions sealed with vacuum grease. Each compartment contained 2 ml of medium. 5-HT 4 agonists were added to the central compartment. Contraction and relaxation of circular muscle were measured in the peripheral orad and caudad compartments, respectively, by using force-displacement transducers attached to the muscle layer. The bathing medium from the central, orad peripheral, and caudad peripheral compartments was removed and stored for measurement of VIP, SP, and CGRP as outlined later. Compartmented Flat-Sheet Preparation of Human Jejunum Segments of normal human jejunum (4 5 cm in length) were obtained from patients undergoing gastric bypass surgery for morbid obesity according to a protocol approved by the Institutional Committee on the Conduct of Human Research. Because the segments of human intestine were small, a two-compartment preparation was used. In some experiments, the 5-HT 4 agonist was added to one compartment and circular muscle response was measured in the other compartment. In other experiments, the order was reversed. The bathing medium from both compartments was removed and stored for measurement of peptide release. Experimental Procedure The experimental procedure was similar in preparations from all three species. After an initial 45-minute equilibration period, a single concentration of 5-HT 4 agonist was added to one compartment for 5 minutes. At the end of this period, the bathing medium from each compartment was removed and stored at 80 C for subsequent immunoassay. Each compartment was then washed with fresh Krebs buffer at 15-minute intervals for 45 minutes, after which another concentration of 5-HT 4 agonist was added. The range of final concentrations of 5-HT 4 agonist in the medium was 0.1 nmol/l to 10 µmol/l. The experiments were repeated in the presence of the selective 5-HT 3 antagonist LY (1-methyl-N-(8-methyl- 8-azabicyclo[3.2.1]-oct-3-yl)-1H-indazole-3-carboxamide maleate); the selective 5-HT 4 antagonist GR A (1-[2- (methylsulfonylamino)ethyl]-4-piperidinyl)methyl 1-methyl- 1H-indole-3-carboxylate, maleate salt); a potent but nonselective 5-HT 4 antagonist SDZ (2-methoxy-4-amino-5-chlorobenzoic acid 2-(diethylamino) ethyl ester); and a CGRP antagonist hcgrp8-37. CGRP release was also measured in the presence of 100 µmol/l hexamethonium. Immunoassays for CGRP, VIP, and SP CGRP was measured by immunoassay as described previously 2 using antibody 6009 raised against human CGRP at a final dilution of 1:24,000. The antibody reacts fully with rat and human CGRP-I and CGRP-II. The limit of detection of the assay was 2.6 fmol/ml, and the median inhibitory concentration (IC 50 ) was 28.6 fmol/ml of original sample. The concentration of CGRP in the samples ranged from 4 to 422 fmol/ml. VIP was measured by immunoassay as described previously 7 using antibody RAS7161 at a final concentration of 1:13,000. The antibody reacts fully with human, rat, and guinea pig VIP. The limit of detection of the assay was 3.0 fmol/ml, and the

3 372 GRIDER ET AL. GASTROENTEROLOGY Vol. 115, No. 2 IC 50 was 18 fmol/ml of original sample. The concentration of VIP in the samples ranged from 5 to 170 fmol/ml. SP was measured by immunoassay as described previously, 12 using antibody RAS7451 at a final concentration of 1:13,000. The antibody reacts fully with SP but not with neurokinin A. The limit of detection of the assay was 3.0 fmol/ml, and the IC 50 was 30 fmol/ml of original sample. The concentration of SP in the samples ranged from 11 to 199 fmol/ml. Data Analysis Contraction and relaxation of circular muscle were measured in grams of force. VIP, SP, and CGRP release were expressed as femtomoles per 100 mg tissue wet weight per minute. The weight of the tissue in the compartments ranged from 112 to 268 mg (mean SE, mg) in guinea pig colon, 184 to 291 mg (mean SE, mg) in rat colon, and 2014 to 3122 mg (mean SE, mg) in human jejunum. Concentrations causing 50% of maximal response (EC 50 ) were calculated from the fit of concentration-response curves using the P.fit 6.0 program (Elsevier, Cambridge). All values are means SE of n experiments where n represents the number of animals. Statistical significance was tested by Student s t test for paired or unpaired values. Materials LY was purchased from Research Biochemicals Inc. (Natick, MA); CGRP, VIP, SP, 125 I-VIP, 125 I-SP, and 125 I-hCGRP from Peninsula Laboratories (Belmont, CA); hc- GRP8-37 from Bachem (Torrance, CA); and bovine serum albumin, phosphoramidon, amastatin and all other chemicals from Sigma Chemical Co. (St. Louis, MO). SDZ and HTF 919 (Zelmac) were a gift of Drs. D. Romer and H.-J. Pfannkuche (Novartis Ltd., Basel, Switzerland). R (Prucalopride) was a gift of Drs. J. Schuurkes and M. Janssens (Janssen Research Foundation, Beerse, Belgium), and GR A was a gift of Drs. G. Kilpatrick and B. Bain (Glaxo Research and Development, Middlesex, England). Results Release of CGRP by the 5-HT 4 Agonist HTF 919 Addition of HTF 919 to the central compartment caused a concentration-dependent release of CGRP into the central compartment but not into the peripheral orad or caudad compartments in rat and guinea pig colon (Figure 1). The EC 50 s for CGRP release into the central compartment were similar in guinea pig ( nmol/l) and rat ( nmol/l) (Table 1), as were the maximal responses elicited by 1 µmol/l HTF 919 ( fmol 100 mg 1 min 1 or 396% 40% above basal release in guinea pig and fmol 100 mg 1 min 1 or 327% 18% above basal release in rat colon). Basal CGRP release was similar in all compartments ranging from to fmol 100 Figure 1. Concentration-response curves for the effect of HTF 919 on (A) CGRP, (B) SP, and (C) VIP release in guinea pig and rat colonic segments. HTF 919 was added to the central compartment, and the release of all three peptides was measured in all compartments (see Materials and Methods for details). The increase in CGRP occurred only in the central compartment, the increase in SP occurred only in the peripheral orad compartment, and the increase in VIP occurred only in the peripheral caudad compartment. Results are expressed as fmol 100 mg 1 min 1. Values are means SE of 4 8 experiments in each species.

4 August HT 4 AGONISTS ELICIT THE PERISTALTIC REFLEX 373 Table 1. EC 50 Values for Release of CGRP, SP, and VIP Induced by the 5-HT 4 Agonist HTF 919 EC 50 (nmol/l) Rat colon Guinea pig colon SP release CGRP release VIP release mg 1 min 1 in guinea pig and to fmol 100 mg 1 min 1 in rat. Release into the peripheral compartments after addition of HTF 919 ( to fmol 100 mg 1 min 1 in guinea pig colon and to fmol 100 mg 1 min 1 in rat colon) was not significantly different from basal release. CGRP release from human intestine in response to 1 µmol/l HTF 919 was fmol 100 mg 1 min 1 above a basal level of fmol 100 mg 1 min 1. When expressed as a percentage of basal release, CGRP release in human intestine (348% 22%) was similar to maximal release in rat and guinea pig colon. CGRP release induced by HTF 919 in all three species was not affected by addition of hexamethonium (100 µmol/l) (Figure 2). Release of SP by HTF 919 The pattern of SP release into the various compartments differed from that of CGRP release. Addition of HTF 919 to the central compartment caused a concentration-dependent release of SP into the peripheral orad compartment but not into the central or peripheral caudad compartments in rat and guinea pig colon Figure 2. CGRP release induced by HTF 919 in humans, guinea pigs, and rats in the presence and absence of hexamethonium. HTF 919 (0.1 µmol/l; ) was added to the central compartment alone or after a 10-minute treatment with hexamethonium (100 µmol/l; ). CGRP release is expressed as fmol 100 mg 1 min 1. Values are means SE of three experiments in each species. (Figure 1). The EC 50 s for SP release into the peripheral orad compartment were similar in guinea pig ( nmol/l) and rat ( nmol/l) (Table 1), as were the maximal responses elicited by 1 µmol/l HTF 919 ( fmol 100 mg 1 min 1 or 183% 9% above basal release in guinea pig and fmol 100 mg 1 min 1 or 198% 5% above basal release in rat colon). Basal SP release was similar in all compartments ranging from to fmol 100 mg 1 min 1 in guinea pig and to fmol 100 mg 1 min 1 in rat. Release into the central and peripheral caudad compartments after addition of HTF 919 ( to fmol 100 mg 1 min 1 in guinea pig colon and to fmol 100 mg 1 min 1 in rat colon) was not significantly different from basal release. SP release from human intestine in response to 1 µmol/l HTF 919 was fmol 100 mg 1 min 1 above a basal level of fmol 100 mg 1 min 1. When expressed as a percentage of basal release, SP release in human intestine (216% 9%) was similar to maximal release in rat and guinea pig colon. Release of VIP by HTF 919 The pattern of VIP release was similar to that of SP release except that it occurred into the peripheral caudad compartment. Addition of HTF 919 to the central compartment caused a concentration-dependent release of VIP into the peripheral caudad compartment but not into the central or peripheral orad compartments in rat and guinea pig colon (Figure 1). EC 50 s for VIP release into the peripheral caudad compartment were similar in guinea pig ( nmol/l) and rat ( nmol/l) (Table 1), as were the maximal responses elicited by 1 µmol/l HTF 919 ( fmol 100 mg 1 min 1 or 203% 3% above basal release in guinea pig and fmol 100 mg 1 min 1 or 193% 6% above basal release in rat colon). Basal SP release was similar in all compartments ranging from to fmol 100 mg 1 min 1 in guinea pig and to fmol 100 mg 1 min 1 in rat. Release into the central and peripheral orad compartments on addition of HTF 919 ( to fmol 100 mg 1 min 1 in guinea pig colon and to fmol 100 mg 1 min 1 in rat colon) was not significantly different from basal release. VIP release from human intestine in response to 1 µmol/l HTF 919 was fmol 100 mg 1 min 1 above a basal level of fmol 100 mg 1 min 1.

5 374 GRIDER ET AL. GASTROENTEROLOGY Vol. 115, No. 2 When expressed as a percentage of basal release, VIP release in human intestine (191% 11%) was similar to maximal release in rat and guinea pig colon. Effect of 5-HT Antagonists on CGRP, SP, and VIP Release Induced by HTF 919 In all three species, addition of the selective 5-HT 3 antagonist LY (10 µmol/l) to the compartment into which HTF 919 was added had no effect on release of CGRP into that compartment, or on release of SP into the peripheral orad compartment, or on release of VIP into the peripheral caudad compartment (Figure 3). In contrast, addition of the mixed 5-HT 4 /5-HT 3 antagonist SDZ (10 µmol/l) virtually abolished (90% 96% inhibition) maximal CGRP, VIP, and SP release induced by 1 µmol/l HTF 919 in all three species (Figure 3). Stimulation of Ascending Contraction and Descending Relaxation of Circular Muscle by HTF 919 Addition of HTF 919 to the central compartment of the three-compartment preparation in guinea pig (Figure 4) and rat colon (Figures 5 and 6) caused a concentration-dependent contraction of circular muscle in the orad compartment (ascending contraction) and relaxation of circular muscle in the caudad compartment (descending relaxation). The EC 50 values for ascending contraction and descending relaxation were closely similar in both species (range, to nmol/l) (Table 2). The maximal responses for ascending contraction and descending relaxation elicited by 1 µmol/l HTF 919 were similar in the two species (Table 2). Addition of the selective 5-HT 4 antagonist GR A or the mixed 5-HT 4 /5-HT 3 antagonist SDZ to the compartment into which HTF 919 was added strongly inhibited ascending contraction and descending relaxation in rat and guinea pig colon (Figures 4 and 5). The responses to the lower concentrations ( 1 nmol/l) were abolished, and the responses to the maximal concentrations were inhibited by 77% 84% in guinea pig colon and by 84% 89% in rat colon. The selective 5-HT 3 antagonist LY had no effect. Ascending contraction and descending relaxation elicited by 1 µmol/l HTF 919 in human jejunum were inhibited 75% 5% and 81% 10%, respectively, by SDZ , to the same extent as maximal responses in guinea pig and rat colon (Figure 7). The selective 5-HT 3 antagonist LY had no effect. Figure 3. Effect of selective 5-HT antagonists on (A) CGRP, (B) SP, and (C) VIP release induced by HTF 919 in segments of human jejunum and rat and guinea pig colon. Sampling of peptides is as detailed in Figure 1. HTF 919 ( ) was added to the mucosa at a maximally effective concentration of 1 µmol/l. Release of all three peptides was not affected by the selective 5-HT 3 antagonist LY ( ) but was virtually abolished by the mixed 5-HT 4 /5-HT 3 antagonist SDZ ( ). Results are expressed as fmol 100 mg 1 min 1. Values are means SE of 3 8 experiments in each species.

6 August HT 4 AGONISTS ELICIT THE PERISTALTIC REFLEX 375 Stimulation of Ascending Contraction and Descending Relaxation of Circular Muscle by R Similar results were obtained with the 5-HT 4 agonist R Addition of R to the central compartment in guinea pig (Figure 8) and rat colon (Figures 8 and 9) caused a concentration-dependent ascending contraction and descending relaxation. The EC 50 values were closely similar for ascending contraction and descending relaxation in rat colon, but were four times lower than the corresponding EC 50 values in guinea pig colon (Table 2). Comparison of the EC 50 s for HTF 919 and R showed that HTF 919 was 4 14 times more potent in rat colon and times more potent in guinea pig colon (Table 2). The maximal responses for ascending contraction and descending relaxation elicited by 1 µmol/l R were similar in the two species Figure 4. Concentration-response curves for the effect of HTF 919 on (A) ascending contraction and (B) descending relaxation in guinea pig colonic segments in the presence and absence of selective 5-HT antagonists. HTF 919 was added to the central compartment; ascending contraction was measured in the peripheral orad compartment, and descending relaxation was measured in the peripheral caudad compartment. Responses were not affected by the selective 5-HT 3 antagonist LY but were inhibited by the selective 5-HT 4 antagonist GR A and the mixed 5-HT 4 /5-HT 3 antagonist SDZ Results are expressed in grams of force. Values are means SE of 3 10 experiments. Figure 5. Concentration-response curves for the effect of HTF 919 on (A) ascending contraction and (B) descending relaxation in rat colonic segments in the presence and absence of selective 5-HT antagonists. HTF 919 was added to the central compartment; ascending contraction was measured in the peripheral orad compartment, and descending relaxation was measured in the peripheral caudad compartment. Responses were not affected by the selective 5-HT 3 antagonist LY , but they were inhibited by the selective 5-HT 4 antagonist GR A and the mixed 5-HT 4 /5-HT 3 antagonist SDZ Results are expressed in grams of force. Values are means SE of 3 8 experiments.

7 376 GRIDER ET AL. GASTROENTEROLOGY Vol. 115, No. 2 Figure 6. Tracings illustrating (A) ascending (orad) contraction and (B) descending (caudad) relaxation of rat colon preparation in response to HTF 919 (1 nmol/l to 1 µmol/l). Asterisks denote addition of HTF 919 to the central compartment. Measurements of response were made in the peripheral orad and caudad compartments. (Table 2), but were significantly lower than the maximal responses elicited by HTF 919. Thus, compared with HTF 919, R was a less potent partial 5-HT 4 agonist. Addition of the selective 5-HT 4 antagonist GR A or the mixed 5-HT 4 /5-HT 3 antagonist SDZ to the compartment into which R was added strongly inhibited ascending contraction and descending relaxation in rat and guinea pig colon (Figures 8 and 9). The responses to the lower concentrations ( 1 nmol/l) were abolished, and the responses to the maximal concentrations were inhibited by 71% 78% in guinea pig colon and by 68% 70% in rat colon. The selective 5-HT 3 antagonist LY had no effect. Effect of the CGRP Antagonist hcgrp8-37 on Ascending Contraction and Descending Relaxation Induced by HTF 919 Addition of 10 µmol/l hcgrp8-37 to the compartment into which HTF 919 was added significantly Table 2. Maximal Responses and EC 50 Values for the Effect of HTF 919 and R on Ascending Contraction and Descending Relaxation inhibited ascending contraction and descending relaxation elicited by 1 µmol/l HTF 919 (Figure 10). Inhibition ranged from 72% 4% to 78% 4% in all three species. Discussion The study shows that selective 5-HT 4 agonists added to the intestinal mucosa trigger the peristaltic reflex: the sequence involves activation of CGRPcontaining sensory nerves resulting in release of excitatory and inhibitory neurotransmitters that regulate the orad and caudad phases of the reflex (Figure 11). CGRP was released only in the compartment where the 5-HT 4 Rat colon Guinea pig colon EC 50 (nmol/l) Response (g) EC 50 (nmol/l) Response (g) HTF 919 Asc. contraction Desc. relaxation R Asc. contraction Desc. relaxation Asc., ascending; Desc., descending. Figure 7. The effect of HTF 919 on (A) ascending contraction and (B) descending relaxation in human jejunal segments in the presence and absence of selective 5-HT antagonists. HTF 919 (1 µmol/l) was added to one compartment of a two-compartment preparation, and ascending contraction or descending relaxation was measured in the other compartment. The responses were not affected by the selective 5-HT 3 antagonist LY but were inhibited by the mixed 5-HT 4 /5- HT 3 antagonist SDZ Results are expressed in grams of force. Values are means SE of three experiments.

8 August HT 4 AGONISTS ELICIT THE PERISTALTIC REFLEX 377 agonist was added, and release was not affected by the presence of hexamethonium; concurrent addition of a CGRP antagonist virtually abolished the ascending and descending phases of the reflex. The pattern implies that the 5-HT 4 receptors that mediate the response to 5-HT 4 agonists are located on sensory CGRP neurons, not on an intermediate cholinergic neuron. Activation of 5-HT 4 receptors located on smooth muscle cells, 27 or cholinergic neurons innervating longitudinal smooth muscle, 28,29 cannot account for the contractile and relaxant phases of the reflex or the sequential release of sensory and motor neurotransmitters. The striking potency of HTF 919, which is effective at subnanomolar concentration in eliciting neurotransmitter release (EC 50, 8 12 nmol/l for CGRP release) and mechanical responses (EC 50, 2 5 nmol/l), is attributable to the proximity of 5-HT 4 receptors located on intrinsic sensory nerve terminals. This suggests that under physi- Figure 8. Concentration-response curves for the effect of RO93877 on (A) ascending contraction and (B) descending relaxation in guinea pig colonic segments in the presence and absence of selective 5-HT antagonists. R was added to the central compartment; ascending contraction was measured in the peripheral orad compartment, and descending relaxation was measured in the peripheral caudad compartment. Responses were not affected by the selective 5-HT 3 antagonist LY , but they were inhibited by the selective 5-HT 4 antagonist GR A and the mixed 5-HT 4 /5-HT 3 antagonist SDZ Results are expressed in grams of force. Values are means SE of 3 6 experiments. Figure 9. Concentration-response curves for the effect of R on (A) ascending contraction and (B) descending relaxation in rat colonic segments in the presence and absence of selective 5-HT antagonists. R was added to the central compartment; ascending contraction was measured in the peripheral orad compartment, and descending relaxation was measured in the peripheral caudad compartment. Responses were not affected by the selective 5-HT 3 antagonist LY , but they were inhibited by the selective 5-HT 4 antagonist GR A and the mixed 5-HT 4 /5-HT 3 antagonist SDZ Results are expressed in grams of force. Values are means SE of 3 8 experiments.

9 378 GRIDER ET AL. GASTROENTEROLOGY Vol. 115, No. 2 of sensory and motor neurotransmitters were similar in both species, consistent with the sequential coupling of sensory and motor neurons. The results using HTF 919 were corroborated by studies using another selective 5-HT 4 agonist, R This agonist also caused concentration-dependent ascending contraction and descending relaxation that were selectively blocked by 5-HT 4 antagonists. Compared with HTF 919, R was less potent, particularly in guinea pig colon, and appeared to be a partial agonist. Consistent with the ability of 5-HT 4 agonists to initiate the peristaltic reflex, application of HTF 919 to the mucosa increased the velocity of propulsion of fecal pellets in an isolated segment of guinea pig colon with a potency (EC 50, 5 nmol/l) 30 similar to that for initiation of peristalsis in flat-sheet compartmented preparation (Tables 1 and 2). Comparative studies showed that R also increased the velocity of propulsion but was about 10 times less potent (EC 50, 50 nmol/l) than HTF 919 and appeared to be a partial agonist. 30 These studies corroborate earlier studies 31,32 of propulsion in guinea pig colonic segments, where 5-HT induced increase in the velocity of propulsion was inhibited by a combination of 5-HT 3 and 5-HT 4 antagonists. In the guinea pig, both 5-HT 3 and 5-HT 4 receptors are involved in sensory neurotransmission and CGRP release. 21 The dual involvement of both 5-HT 4 and 5-HT 3 receptors is evident when 5-HT Figure 10. Effect of the selective CGRP antagonist hcgrp8-37 on (A) ascending contraction and (B) descending relaxation elicited by HTF 919 in segments of human jejunum and rat and guinea pig colon., HTF 919; J, hcgrp8-37. HTF was added to the mucosa at a maximal concentration of 1 µmol/l. The responses in all three species were inhibited by 75% 80% by hcgrp8-37 (10 µmol/l). Results were expressed in grams of force. Values are means SE of 3 4 experiments in each species. ological conditions, minimal amounts of 5-HT released from mucosal enterochromaffin cells may be sufficient to activate 5-HT 4 receptors on sensory neurons and thus trigger a peristaltic reflex. Our previous studies 20,21 had shown that 5-HT added to the mucosal surface elicits release of the sensory neurotransmitter CGRP. Although VIP and SP are not the sole inhibitory or excitatory neurotransmitters released during peristalsis, they were measured in this study as markers of transmitter release that occurs during the descending and ascending phases of the reflex. The EC 50 s for HTF 919 that elicited release Figure 11. Model depicting extrinsic and intrinsic sensory pathways mediating the peristaltic reflex. The extrinsic pathway is activated by muscle stretch and consists of CGRP neurons with cell bodies in the dorsal root ganglion. The intrinsic pathway is activated by mucosal stimuli and consists of CGRP neurons with cell bodies in the enteric nervous system and afferent projections to the mucosa. Mucosal stimuli release 5-HT from enterochromaffin cells that acts on 5-HT 4 receptors (in humans and rats) and both 5-HT 4 and 5-HT 3 receptors (in guinea pigs) located on the nerve terminals of intrinsic CGRP neurons. Whether extrinsic or intrinsic, CGRP neurons relay sensory stimuli to the same populations of interneurons coupled to excitatory cholinergic and tachykinin motor neurons and inhibitory VIP, PACAP, and NO synthase motor neurons.

10 August HT 4 AGONISTS ELICIT THE PERISTALTIC REFLEX 379 is added exogenously or released endogenously from enterochromaffin cells. Previous studies 33 using the Trendelenburg preparation, which provides an overall index of peristaltic activity but does not distinguish the phases of the peristaltic reflex, had shown that HTF 919 potently stimulates peristaltic activity in guinea pig intestinal segments with an EC 50 of 20 nmol/l. In vivo, HTF 919 restored colonic motor function in lidamidine-treated mice, 33 and increased colonic transit in humans and dogs. 34,35 R exhibited similar properties, increasing colonic transit and the frequency of defecation in humans, dogs, and cats. 26,36,37 Studies using muscle strips from various species suggested that R was a partial 5-HT 4 agonist with a potency (EC 50, nmol/l) 26 similar to that reported in the present study. The effectiveness of 5-HT 4 agonists when applied to the intestinal mucosa underscores their potential as therapeutic agents. Low concentrations of these agents can trigger a physiological reflex that enhances colonic propulsive activity, thereby decreasing the likelihood of side effects by obviating the need for distribution in the body. References 1. Grider JR, Jin J-G. Distinct populations of sensory neurons mediate the peristaltic reflex elicited by muscle stretch and mucosal stimulation. J Neurosci 1994;14: Grider JR. CGRP as a transmitter in the sensory pathway mediating the peristaltic reflex. Am J Physiol 1994;266:G1139 G Smith TK, Bornstein JC, Furness JB. Interactions between reflexes evoked by distension and mucosal stimulation: electrophysiological studies of guinea-pig ileum. J Auton Nerv Syst 1991;34: Smith TK, Bornstein JC, Furness JB. Convergence of reflex pathways excited by distension and mechanical stimulation of the mucosa onto the same neurons of the guinea pig small intestine. J Neurosci 1992;12: Grider JR. Identification of neurotransmitters regulating reciprocal longitudinal muscle responses during the peristaltic reflex (abstr). Gastroenterology 1995;108:A Sarna SK. Gastrointestinal longitudinal muscle contractions. Am J Physiol 1993;265:G156 G Grider JR, Makhlouf GM. Colonic peristaltic reflex: identification of VIP as a mediator of descending relaxation. Am J Physiol 1986; 251:G40 G Grider JR, Katsoulis S, Schmidt WE, Jin J-G. Regulation of the descending relaxation phase of intestinal peristalsis by PACAP. J Auton Nerv Syst 1994;50: Grider JR. Interplay of VIP and nitric oxide in the regulation of the descending relaxation phase of peristalsis. Am J Physiol 1993; 264:G334 G Makhlouf GM, Grider JR. Nonadrenergic, noncholinergic inhibitory transmitters of the gut. News Physiol Sci 1993;8: Grider JR. Tachykinins as transmitters of the ascending contractile component of the peristaltic reflex. Am J Physiol 1989;257: G709 G Holzer P, Schluet W, Maggi CA. Ascending enteric reflex contraction: roles of acetylcholine and tachykinins in relation to distension and propagation of excitation. J Pharmacol Exp Ther 1993; 264: Giuliani S, Lecci A, Giachetti A, Maggi CA. Tachykinins and reflexly evoked atropine-resistant motility in the guinea pig colon in vivo. J Pharmacol Exp Ther 1993;265: Bornstein JC, Furness JB, Smith TK, Trussell DC. Synaptic responses evoked by mechanical stimulation of the mucosa in morphologically characterized myenteric neurons of the guineapig ileum. J Neurosci 1991;11: Yuan SY, Furness JB, Bornstein JC. Post-stimulus depression of reflex changes in circular muscle activity in the guinea pig small intestine. J Auton Nerv Syst 1992;40: Yuan SY, Furness JB, Bornstein JC, Smith TK. Mucosal distortion by compression elicits polarized reflexes and enhances responses of the circular muscle to distension in the small intestine. J Auton Nerv Syst 1991;25: Grider JR. Identification of neurotransmitter regulating intestinal peristaltic reflex in humans. Gastroenterology 1989;97: Costa M, Furness JB. The peristaltic reflex: an analysis of nerve pathways and their pharmacology. Naunyn Schmiedebergs Arch Pharmacol 1976;294: Smith TK, Bornstein JC, Furness JB. Distension-evoked ascending and descending reflexes in the circular muscle of guinea-pig ileum: an intracellular study. J Auton Nerv Syst 1990;29: Grider JR, Kuemmerle JF, Jin J-G. 5-HT released by mucosal stimuli initiates peristalsis by activating 5-HT 4 /5-HT 1p receptors on sensory CGRP neurons. Am J Physiol 1996;270:G778 G Foxx-Orenstein AE, Kuemmerle JF, Grider JR. Distinct 5-HT receptors mediate the peristaltic reflex induced by mucosal stimuli in human and guinea pig intestine. Gastroenterology 1996;111: Furness JB, Johnson PJ, Pompolo S, Bornstein JC. Evidence that enteric motility reflexes can be initiated through entirely intrinsic mechanisms in the guinea-pig small intestine. Neurogastroenterol Motil 1995;7: Pan H, Wade PR, Gershon MD. Activation of and transmission from intrinsic sensory neurons of the submucosal plexus (abstr). Gastroenterology 1997;112:A Buchheit K-H, Gamse R, Giger R, Hoyer D, Klein F, Kloppner E, Pfannkuche H-J, Mattes H. The serotonin 5-HT 4 receptor. 1. Design of a new class of agonists and receptor map of the agonists recognition site. J Med Chem 1995;38: Buchheit K-H, Gamse R, Giger R, Hoyer D, Klein F, Kloppner E, Pfannkuche H-J, Mattes H. The serotonin 5-HT 4 receptor. 2. Structure-activity studies of the indole carbazimidamide class of agonists. J Med Chem 1995;38: Briejer MR, Meulemansd AL, Bosmans J-P, Van Daele P, Schuurkes JAJ. In vitro pharmacology of the novel enterokinetic R (abstr). Gastroenterology 1997;112:A Kuemmerle JF, Murthy KS, Grider JR, Martin DC, Makhlouf GM. Coexpression of 5-HT 2A and 5-HT 4 receptors coupled to distinct signaling pathways in human intestinal muscles cells. Gastroenterology 1995;109: Kilbinger H, Wolf D. Effects of 5-HT 4 receptor stimulation on basal and electrically evoked release of acetylcholine from guinea-pig myenteric plexus. Naunyn Schmiedebergs Arch Pharmacol 1992; 345: Tonini M, Galligan JJ, North RR. Effects of cisapride on cholinergic neurotransmission and propulsive motility in the guinea pig ileum. Gastroenterology 1989;96: Jin JG, Foxx-Orenstein AE, Grider JR. Stimulation of colonic propulsion by 5-HT 4 receptor agonists: synergism by opioid receptor antagonists (abstr). Gastroenterology 1997;112:A Wade PR, Chen J, Jaffe B, Kassem IS, Blakely RD, Gershon MD.

11 380 GRIDER ET AL. GASTROENTEROLOGY Vol. 115, No. 2 Localization and function of a 5-HT transporter in crypt epithelia of the gastrointestinal tract. J Neurosci 1996;16: Kadowaki M, Wade PR, Gershon MD. Participation of 5-HT 3, 5-HT 4 and nicotinic receptors in the peristaltic reflex of guinea pig distal colon. Am J Physiol 1996;271:G849 G Pfannkuche H-J, Buhl T, Gamse R, Hoyer D, Mattes H, Buchheit K-H. The properties of a new prokinetically active drug, SDZ HTF 919 (abstr). Neurogastroenterol Motil 1995;7: Appel S, Kumle A, Hubert M, Duvauchelle T. First pharmacokineticpharmacodynamic study in humans with a selective 5-hydroxytryptamine 4 receptor agonist. J Clin Pharmacol 1997;37: Nguyen A, Camilleri M, Kost LJ, Metzger A, Sarr MG, Hanson RB, Fett SL, Zinsmeister AR. SDZ HTF 919 stimulates canine colonic motility and transit in vivo. J Pharmacol Exp Ther 1997;180: Briejer MR, Van Daele P, Bosmans J-P, Ghoos E, Eelen J, Schuurkes JAJ. Dose-dependent effects after oral and intravenous administration of R on colonic motility in conscious dogs (abstr). Gastroenterology 1997;112:A Briejer MR, Engelen M, Jacobs J, Vlaminck K, Schuurkes JAJ. R enhances defecation frequency in conscious cats (abstr). Gastroenterology 1997;112:A705. Received September 29, Accepted May 5, Address requests for reprints to: John R. Grider, Ph.D., Box , Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia jgrider@gems.vcu.edu; fax: (804) Supported by grant DK34153 from the National Institute of Diabetes and Digestive and Kidney Diseases.