- Email: [email protected]
Identification of neurotransmitters regulating intestinal peristaltic reflex in humans
GASTROENTEROLOGY
1989;97:1414-9
Identification of Neurotransmitters Regulating Intestinal Peristaltic Reflex in Humans J. R. GRIDER Departments
of Medicine
and Physiology,
Medical College of Virginia, Richmond,
The components of the intestinal peristaltic reflex in humans were examined and the neurotransmitters responsible for them identified for the first time in isolated flat sheet segments of intestine. Increasing radial stretch to the caudad end elicited increasing ascending contraction only, whereas increasing radial stretch to the orad end elicited increasing descending relaxation only. Both components were abolished by hexamethonium, implying the participation of cholinergic interneurons in each component. Atropine inhibited ascending contraction only, abolishing the response to low grades of stretch and partially inhibiting the response to high grades of stretch (69% f 17% p < 0.01). The substance P antagonist [D-Pro2, D-Trp7Yg]substance P partially inhibited ascending contraction induced by high grades of stretch only (40% + 12%, p < 0.02). The vasoactive intestinal peptide antagonist [4-Cl-D-Phe’, Leul’]vasoacfive intestinal peptide inhibited descending relaxation, abolishing the response to low grades of stretch and partially inhibiting the response to high grades of stretch (40% f 4%, p < 0.001). Release of vasoactive intestinal peptide increased significantly by 91% during descending relaxation only, whereas release of both substance P and substance K increased significantly by 107% during ascending contraction only, supporting the participation of vasoactive intestinal peptide motor neurons in descending relaxation and tachykinin motor neurons as well as cholinergic motor neurons in ascending contraction. The components of the human peristaltic reflex and transmitters regulating them were identical to those found in rat and guinea pig intestine. n human and other mammalian species, the myenteric plexus of the enteric nervous system consists of two main populations of neurons that can be distinguished by their content of peptide transmitter. The neurons contain either vasoactive intestinal
I
Virginia
peptide (VIP) and its homologue, peptide histidine methionine (PHM in humans), OI substance P (SP) and its homologue substance K (SK), also known as neurokinin A (1,~). Studies on rat and guinea pig intestine indicate that VIP motor neurons regulate the descending relaxation component of the peristaltic reflex, whereas SP motor neurons regulate in part the ascending contraction component of the reflex (3). Evidence for VIP as the relaxant transmitter in myenteric motor neurons is based on the following: 1. Vasoactive intestinal peptide and its homologues have the unique ability to cause relaxation in circular smooth muscle throughout the digestive tract (4-10). 2. Vasoactive intestinal peptide is released during electrical or reflex activation of myenteric neurons and its release is proportional to the intensity of stimulation and accompanied by a corresponding increase in relaxation (11). induced relaxation is selectively 3. Neurally blocked by specific VIP antiserum and selective VIP antagonists (3,4,6,7,11-17). Evidence for SP [and substance contractile transmitter in myenteric is based on the following:
K (SK]] as a motor neurons
1. Substance P and SK cause direct and cholinergically mediated contraction (18-21). 2. Substance P is released during electrical or reflex activation of myenteric neurons (22-25).
Abbreviations used in this paper: NKA, neurokinin A; PHI, peptide histidine isoleucine; PHM, peptide histidine methionine; SK, substance K; SP, substance P. 0 1989 by the American Gastroenterological Association QQl8-5085/89/$3.59
December
HUMAN PERISTALTIC REFLEX
1989
3. Neurally induced by SP antagonists
contraction is partially blocked (3,20,21,26).
In recent studies on guinea pig and rat (3)) we have used a preparation consisting of an isolated whole segment of small intestine or colon that enables characterization of the components of the peristaltic reflex (ascending contraction and descending relaxation) separately. In the present study, we have modified this preparation to enable recording of the components of the peristaltic reflex from flat sheets of human intestine. The results show that the components of the peristaltic reflex in humans are identical to those in rat and guinea pig. Vasoactive intestinal peptide is released during and is responsible for descending relaxation, whereas acetylcholine is responsible for ascending contraction induced by low grades of stretch and tachykinins are responsible for ascending contraction induced by high grades of stretch.
Materials and Methods Preparation
of Flat
Intestinal
Sheets
A single segment of human jejunum was obtained from each of 22 patients undergoing bypass surgery for morbid obesity; the segments were labeled with a stitch to indicate oral and caudad ends, transported in ice-cold saline, and prepared for the experiments within 30 min of excision. The segments were opened along the mesenteric border and the mucosa was removed by blunt dissection. The segments were on average 6 cm long and 3 cm wide in the unstretched state. The segments were pinned flat on a base of sylgard in an organ bath (Figure 1) containing 10 ml of Krebs’ bicarbonate solution with the following composition (in mM): 118 NaCl, 4.8 KCl, 1.2 KH,PO,, 1.2 MgSO,, 2.5 CaCl,, 25 NaHCO,, 11 glucose, and 0.1% bovine serum albumin. The solution was maintained at 37°C and gassed with 95% O,-5% CO,.
Figure 1 Flat sheet preparation of human intestine for measurement of components of peristaltic reflex. Hook-andpulley assemblies permit radial stretch of circular muscle by application of variable weights (W) to one end and measurement of circular muscle response via a force-displacement transducer at the other end.
Experimental
1415
Procedure
The segments were immobilized so as to enable radial stretch of one end of the segment and recording from the other end using a force displacement transducer (Grass FT03C). Hook-and-pulley assemblies were used to apply radial stretch to and record from the horizontal sheet (27). A set of 4-5 pins secured the midportion
of the segment,
thus separating the stretch site from the recording site. The distance between the stretch site and the recording site was about 4 cm. Radial stretch was applied in the range of 2-14 g at the extreme orad end to elicit descending relaxation only and at the extreme caudad end to elicit ascending contraction only.
Experimental
Design
The jejunal segment was incubated for 1 h, during which the Krebs’ bicarbonate solution was changed four times. The preparation developed basal tone (0.5 g) within 30 min, which remained stable thereafter.
Control stretch-
response curves were then constructed by applying weights in 2-g increments in the range of 2-14 g to the hook-and-pulley assembly. Each increment was applied for a period of 15 s at intervals of 2 min. After 1 h to allow for recovery of the preparation, antagonists were added for 15 min and another stretch-response curve was constructed. The antagonists used included the muscarinic antagonist 1 PM atropine, the nicotinic antagonist 0.1 mM hexamethonium, the VIP/peptide histidine isoleucine (PHI) antagonist 10 PM [&Cl-D-Phe’, Leu17]VIP, and the SP/SK antagonist 10 PM [D-Pro’, D-Trp7*9]SP. The antagonists were used at maximally effective concentrations as determined in previous studies (3,28). Neither the antagonists nor the axonal blocker tetrodotoxin had any effect on basal tone. At the completion of the experiments with antagonists, the bathing medium was changed four times and the preparation retested; in all experiments, the preparation reverted to control responses. For measurement of peptide release in some experiments, the bathing solution was collected during a 15-min basal period and during a subsequent 15-min period in which either ascending contraction or descending relaxation was induced as described by applying increasing stretch for 15 s at intervals of 2 min in the range of 2-12 g. Samples were collected on ice in plastic tubes containing bacitracin (final concentration 20 PM) and aprotinin (1000 U/ml). The samples were divided into l-ml aliquots and stored at - 20°C for subsequent radioimmunoassay. Vasoactive intestinal peptide was measured in duplicate using antiserum AC115 (final concentration 1: 50,000) as described (6,7). The limit of detection of the assay was 5 pg/ml of original sample. The range of concentrations in all samples was 35-220 pg/ml. Substance K was measured in duplicate using SK antiserum 7359N (final concentration 1:2500). The antiserum was specific for SK in that SP cross-reacted only minimally with it (~3%); other peptides such as bombesin and neurotensin did not crossreact. The limit of detection of the assay was 5 pg/ml. The range of concentrations of SK in all samples was 7-105 pg/ml. Substance P was also measured in duplicate using a
1416
GASTROENTEROLOGY Vol. 97, No. 6
GRIDER
ASCENDING
specific SP antiserum, AC95 (final concentration 1:2,500). Substance K, bombesin, neuromedin K, VIP, PHI, dynorphin-13, and somatostatin did not cross-react with this antiserum. The limit of detection of the assay was 10 pg/ml and the IC,, was 130 pg/ml.
Data
DESCENDING
CONTRACTION
RELAXATION
i
Analysis
Descending relaxation and ascending contraction were expressed as percentages of the maximal response obtained using a stretch stimulus of 12 g. Statistical significance was evaluated using Student’s t-test for paired values. Only one experiment was done in each segment. Each experiment consisted of a control measurement in the absence and another in the presence of a specific antagonist; n designates the number of experiments.
Materials [D-Pro’, D-Trp’,‘]SP and [4-Cl-D-Phe’, Leu”lVIP were obtained from Bachem, Torrance, Calif.; VIP antiserum AC115 from Cambridge Research Biochemicals, Valley Stream, N.Y.; SK antiserum 7359N and ?-SK from Peninsula Laboratories, Belmont, Calif.; ‘*‘I-VIP from New England Nuclear, Boston, Mass.; aprotinin from Mobay Chemicals, New York, N.Y.; and bacitracin, atropine, hexamethonium, and bovine serum albumin from Sigma Chemicals, St Louis, MO.
/;
20
I
OL
,
I
I
0
4
CAUDAD
I
I
I
8
STRETCH
1
,
12
(q 1
c
,
,
,
0
4
,
(
, 8
I2
ORAD STRETCH
lq)
Figure 3. Left panel: ascending contraction in response to increasing radial stretch of the caudad end of the human intestinal segment. Right panel: descending relaxation in response to increasing radial stretch of the orad end. Data are expressed as a percentage of maximal response. Values are mean t SE of 7-11 experiments.
ascending contraction response was 0.9 ? 0.1 g (n = 81, and the maximal descending relaxation response was 0.4 ? 0.06 g [n = 11). The pattern was identical to that obtained in whole segments of guinea pig and rat colon (3). Peptide Release During Ascending Contraction and Descending Relaxation
Results Components Reflex
of the Intestinal
Peristaltic
Release of VIP during a 15-min basal period was 96 + 24 pg/g tissue wet weight and increased by
Typical tracings of ascending contraction and descending relaxation are illustrated in Figure 2. Graded radial stretch of the orad end of a flat sheet intestinal preparation elicited increasing descending relaxation only (Figure 3, right). Increasing radial stretch of the caudad end elicited increasing ascending contraction only [Figure 3, left). The maximal
ASCENDING CONTRACTION
DESCENDING RELAXATION
it
T
MD120 -
eo -
b _ h _ - _ - -l_
-h-b-h-
10 P STRETCH
-i
H
;
-+
q-
60 IIJi %3 w J= LlJO aL
4O20 -
/o.59
Figure 2. Upper panel: typical ascending contraction response elicited by radial stretch of the caudad end of the human intestinal segment. Lower panel: typical descending relaxation response elicited by radial stretch of the caudad end of the human intestinal segment. Numbers in the center indicate the grams of stretch applied via the hook-and-pulley assembly. The horizontal bars indicate the period of stretch (15 s).For the remainder of the 2-min period before application of the next stimulus, the preparation reverted to basal level.
-60
t
1
Figure 4. Release of VIP, SK, and SP during ascending contraction and descending relaxation. Data are expressed as a percentage change from basal release. Values are mean 2 SE of 3-6 experiments.
December 1989
HUMAN PERISTALTIC REFLEX
ASCENDING CONTRACTION
ASCENDING CONTRACTION
DESCENDING RELAXATION
VIP
4
0 CAUDAD
Figure
6 STRETCH
DESCENDING RELAXATION
ANTAGONIST
12
VIP ANTAGONIST
((1)
ORAD
STRO%H
(g)
5. Left panel: ascending contraction in response to increasing radial stretch of the caudad end of human intestinal segment in the absence (closed circles] and presence (open circles) of atropine (1 PM). Right panel: descending relaxation in response to increasing radial stretch of the orad end in the absence [closed circles] and presence (open circles) of atropine (1 PM). Data are expressed as a percentage of maximal response. Values are mean 2 SE of 4 experiments.
91% k 28% (p < 0.02;n = 6)during a 15-min period in which descending relaxation was induced by progressive radial stretch of the extreme orad end (Figure 4). In contrast, release of VIP decreased by 31% + 23% (NS;n = 3) during a 15-min period in which ascending contraction was induced by progressive radial stretch of the extreme caudad end (Figure 4). The pattern of release of SK was the reverse of that for VIP (Figure 4). Release of SK during a 15-min basal period was 59 ? 8 pg/g tissue wet weight and increased by 107% + 32% (p < 0.05) during a 15-min period in which ascending contraction was induced by progressive radial stretch of the extreme caudad 100
1417
CONTROL
HUMAN JEJUNUM
SP ANTAGONIST
ATROPINE
OL
0
1
1
4
CAUDAD Figure
7.
*
j
8
STRETCH
j
c
12 (g)
1
0
4 ORAD
6 STRETCH
12 (g 1
Effect of VIP antagonist {B-Cl-D-Phe’, Leu171VIP (10 PM)} on ascending contrktion (left panel) kd descending relaxation (right panel]. Data are expressed as a percentage of maximal response. Values are mean 2 SE of 3-5 experiments.
end. In contrast, release of SK decreased by 12% + 9% (NS) during a 15min period in which descending relaxation was induced by progressive radial stretch of the extreme orad end. The pattern of release of SP was identical to that of SK. During the 15min period in which ascending contraction was induced, release of SP increased by 108% 2 46% above basal level (p < 0.05), whereas during the 15-min period in which descending relaxation was induced, SP release was decreased by 14% + 17% (NS] Efiects of Nicotinic and Muscarinic Antagonists on the Components of the Peristaltic Reflex Tetrodotoxin (1 PM) and hexamethonium (100 PM) abolished both ascending contraction and descending relaxation at all grades of stretch. Atropine (1 PM) had no effect on descending relaxation but inhibited ascending contraction at all grades of stretch (Figure 5). The ascending contraction response to 2- and 4-g stimuli was abolished, whereas the response to 10 g was inhibited by 69% 2 17% (p < 0.01; n = 4).
ANTA%NlST ATR;PlNE 0
4 CAUDAD
Figure
6.
8 STRETCH
12 (g)
Effect of atropine (1 PM), SP antagonist {[D-Pro’, DTrp7,‘]SP (10 j&f)}, and a combination of atropine and SP antagonist on ascending contraction induced by various grades of stretch. Data are expressed as a percentage of maximal response. Values are mean 2 SE of 4 experiments.
Effect of Tachykinin Antagonist on the Components of the Peristaltic Reflex The tachykinin antagonist [D-Pro’, D-Trp’,‘]SP (10 PM), like atropine, had no effect on descending relaxation, but inhibited ascending contraction (Figure 6).However, unlike atropine, it had little or no effect on ascending contraction induced by the lowest grades of stretch (2 and 4 g), but inhibited
1418
GASTROENTEROLOGY
GRIDER
ascending contraction induced by the higher grades of stretch; the response to 12 g of stretch was inhibited by 40% -+12% (p < 0.02;n = 4) (Figure 6).The combination of atropine and tachykinin antagonist abolished the response to low grades of stretch and inhibited the response to higher grades of stretch by 84% + 6% (p < 0.01;n = 4).After removal of the antagonists from the bathing medium by repeated washing, the responses returned to control levels. Effect of Vasoactive Intestinal Peptide Antagonist on the Components of the Peristaltic Reflex The VIP antagonist [4Cl-D-Phe”, Leu’7]VIP (10 PM) inhibited descending relaxation at all grades of stretch (Figure 7). The response to the lowest grades of stretch (2 and 4 g) was abolished and the response to the highest grade of stretch (12 g) was inhibited by 40% + 4% (p < 0.001;n = 5). In contrast, the VIP antagonist augmented ascending contraction at all grades of stretch. The response to the lowest grade of stretch (2g)was increased by 79% + 4% (p < 0.01, n = 3) and the response to the highest grade of stretch was increased by 19% t 1% (p < 0.01;n = 3).
Discussion This study identifies for the first time the peptide transmitters involved in the regulation of the intestinal peristaltic reflex in humans. For this purpose, small segments of human intestine obtained at surgery were used as flat sheets to examine separately the components of the peristaltic reflex. The components of the reflex (ascending contraction and descending relaxation) were shown to be identical to those in the colon of the rat and guinea pig (3). The identity of the peristaltic reflex in humans and animals extends to the following. First, in both human and animal intestine (3,27), the pathways regulating ascending contraction and descending relaxation involve the participation of cholinergic interneurons, as both components are abolished by nicotinic antagonists. Second, cholinergic motor neurons are involved in the regulation of ascending contraction only (Figure 5). Ascending contraction induced by low grades of stretch is abolished by atropine, implying that at this intensity of stimulation cholinergic motor neurons only are involved. Third, tachykinin neurons are involved in the regulation of ascending contraction induced by high grades of stretch; at this intensity of stimulation, both tachykinin and cholinergic neurons are involved because ascending contraction induced by high grades of stretch is sensitive to both atropine
Vol. 97, No. 6
and SP antagonists (Figure 6).In human as in animal intestine, ascending contraction only is accompanied by SK and SP release. The release of both SP and SK implies that the tachykinin precursor in these motor neurons is /?-preprotachykinin which contains the sequences of both peptides. Fourth, VIP motor neurons regulate descending relaxation. In human and animal intestine (3), descending relaxation only is accompanied by release of VIP. The VIP release is coupled to descending relaxation because selective VIP antagonists inhibit descending relaxation (28).In animal intestine, in addition, descending relaxation is selectively inhibited by VIP-specific antiserum in a concentrationdependent manner (3). It is noteworthy that descending relaxation induced by low grades of stretch in human and animal intestine is abolished by the VIP antagonist, whereas descending relaxation induced by high grades of stretch is partially inhibited consistent with competitive antagonism. As shown previously (28),the antagonist used in the present study inhibited relaxation induced by both VIP and PHI. The extent of the contribution of PHM (or PHI) is not known, but is likely to be minor as PHUPHM, though derived from the same precursor as VIP, is more extensively processed in tissues and is less potent (5-10 times) than VIP as a relaxant agent (29,30). The augmentation of ascending contraction by VIP antagonist, which is also observed with VIP antiserum in animal studies, implies that background VIP influences the contractile state and response of circular smooth muscle. Recent studies support the notion that the inhibitory activity of VIP neurons masks myogenic phasic activity, which can be restored by addition of VIP antiserum or VIP antagonists (31). Thus, the influence of cholinergic and tachykinin neurons in the regulation of ascending contraction should be viewed in relation to the background influence of VIP neurons.
References Wattchow DA, Furness JB, Costa M. Distribution and coexistence of peptides in nerve fibers of the external muscle of the human gastrointestinal tract. Gastroenterology 1988;95:3241. Llewellyn-Smith IJ, Furness JB, Costa M, Gibbins IL. Quantitative ultrastructural analysis of enkephalin-, substance Pand VIP-immunoreactive nerve fibers in the circular muscle of the guinea pig small intestine. J Comp Neurol 1988; 272:139-48. Grider JR, Makhlouf GM. Colonic peristaltic reflex: identification of vasoactive intestinal peptide as mediator of descending relaxation. Am J Physiol 1986;251:G40-5. Biancani P, Walsh JH, Behar J. Vasoactive intestinal peptide: a neurotransmitter for relaxation of the rabbit internal anal sphincter. Gastroenterology 1985;89:867-74.
December 1989
5. Bitar KN, Makhlouf GM. Relaxation of isolated gastric smooth muscle cells by vasoactive intestinal peptide. Science (Wash DC) 1982:261:531-3. 6. Grider JR, Cable MB, Said SI, Makhlouf GM. Vasoactive intestinal peptide (VIP] as neural mediator of gastric relaxation. Am J Physiol 1985;248:G73-8. 7. Grider JR, Cable MB, Bitar KN, Said SI, Makhlouf GM. Vasoactive intestinal peptide. Relaxant neurotransmitter in tenia coli of the guinea pig. Gastroenterology 1985;89:36-42. 8. Makhlouf GM. Enteric neuropeptides: role in neuromuscular activity of the gut. Trends Pharmacol Sci 1985;6:214-8. 9. Larsson LT, Malmfors G, Wahlestedt C, Leander S, Hakanson R. Hirschprung’s disease: a comparison of the nervous control of ganglionic and aganglionic smooth muscle in vitro. J Pediatr Surg 1987;22:431-5. 10. Furness JB, Costa M. VIP and enteric inhibitory nerves. In: Said S, ed. Vasoactive intestinal peptide. New York: Raven, 1982:391-406. 11. Grider JR, Makhlouf GM. Prejunctional inhibition of vasoactive intestinal peptide release. Am J Physiol 1987;253:G7-12. 12. Angel F, Go VLW, Schmalz PF, Szurszewski JH. Vasoactive intestinal polypeptide. A putative neurotransmitter in the canine gastric muscularis mucosae. J Physiol Lond 1983; 341:641-5. 13. Biancani P, Walsh JH, Behar J. Vasbactive intestinal polypeptide. A neurotransmitter for lower esophageal sphincter relaxation. J Clin Invest 1984;73:963-7. 14. D’Amato M, DeBeurme FA, Lefebvre A. Comparison of the effect of vasoactive intestinal polypeptide and non-adrenergic, non-cholinergic neurone stimulation in the cat gastric fundus. Eur J Pharmacol 1988;152:71-82. 15. Goyal RK, Rattan S, Said SI. VIP as a possible neurotransmitter of non-cholinergic, non-adrenergic inhibitory neurons. Nature 1985;288:378-80. 16. Nurko S, Rattan S. Role of vasoactive intestinal polypeptide in the internal anal sphincter relaxation of the opossum. J Clin Invest 1988;81:1146-53. 17. Wiley JW, O’Dorisio TM, Owyang C. Vasoactive intestinal polypeptide mediates cholecystokinin-induced relaxation of the sphincter of Oddi. J Clin Invest 1988:81:1920-l. 18 Souquet JC, Bitar KN, Grider JR, Makhlouf GM. Receptors for substance P on isolated intestinal smooth muscle cells of the guinea pig. Am J Physiol 1987;253:G666-72. 19. Fosbraey P, Featherstone RL, Morton IKM. Comparison of potency of substance P and related peptides on 13H]acetylcholine release and contractile actions in the guinea pig ileum. Naunyn Schmiedebergs Arch Pharmacol 1984;326: 111-5. 20 Kilbinger H, Staub P, Erlhof I, Holzer P. Antagonist discimi-
HUMAN PERISTALTIC REFLEX
1419
nation between subtypes of tachykinin receptors in the guinea pig. Naunyn Schmiedebergs Arch Pharmacol 1986; 334:181-7. 21. Costa M, Furness JB, Pullin CO, Bornstein J. Substance P enteric neurons mediate non-cholinergic transmission to the circular muscle of the guinea pig intestine. Naunyn Schmiedebergs Arch Pharmacol 1985;328:446-63. 22. Donnerer J, Barth0 L, Holzer P, Lembeck F. Intestinal peristalsis associated with the release of immunoreactive substance P. Neuroscience 1984;11:913-18. 23. Barson SA, Jaffe BM, Gintzler AR. Release of substance P from the enteric nervous system: direct quantitation and characterization. J Pharmacol Expt Ther 1983;227:365-8. 24. Grider JR, Makhlouf GM. Regulation of the ascending contraction component of the peristaltic reflex by myenteric tachykinin neurons (abstr). Gastroenterology 1988;94:A157. release of 25. Holzer P. Characterization of the stimulus-induced immunoreactive substance P from the myenteric plexus of the guinea pig small intestine. Brain Res 1984;297:127-36. 26. Barth0 L, Holzer P, Donnerer J, Lembeck F. Effects of substance P, cholecystokinin octapeptide, bombesin, neurotensin on the peristaltic reflex of the guinea pig ileum in the absence and presence of atropine. Naunyn Schmiedebergs Arch Pharmacol 1982;321:321-8. 27. Costa M, Furness JB. The peristaltic reflex: an analysis of the nerve pathways and their pharmacology. Naunyn Schmiedebergs Arch Pharmacol 1976;294:47-60. 28. Grider JR, Rivier J, Makhlouf GM. Evidence for VIP as transmitter of inhibitory motor neurons: blockade of neurally mediated relaxation by VIP antagonists (abstr). Gastroenterology 1987;92:1415. 29. Fahrenkrug J, Bek T, Lundberg JM, Hokfelt T. VIP and PHI in cat neurons: colocalization but variable tissue content possibly due to differential processing. Regul Pept 1985;12:21-34. 30. Itoh N, Obata K, Yanihara N, Okamoto H. Human preprovasoactive intestinal peptide contains a novel PHI-27-like peptide, PHM-27. Nature 1983;304:547-9. 31 Grider JR, Makhlouf GM. Suppression of inhibitory neural input to colonic circular muscle by opioid peptides. J Pharmacol Exp Ther 1987;243:205-10.
Received January 6, 1989. Accepted May 31,1989. Address requests for reprints to: J. R. Grider, Ph.D., Box 711, MCV Station, Medical College of Virginia, Richmond, Virginia 23298-711. This work was supported by grant DK34153 from the National Institute of Diabetes, and Digestive and Kidney Diseases.
1989;97:1414-9
Identification of Neurotransmitters Regulating Intestinal Peristaltic Reflex in Humans J. R. GRIDER Departments
of Medicine
and Physiology,
Medical College of Virginia, Richmond,
The components of the intestinal peristaltic reflex in humans were examined and the neurotransmitters responsible for them identified for the first time in isolated flat sheet segments of intestine. Increasing radial stretch to the caudad end elicited increasing ascending contraction only, whereas increasing radial stretch to the orad end elicited increasing descending relaxation only. Both components were abolished by hexamethonium, implying the participation of cholinergic interneurons in each component. Atropine inhibited ascending contraction only, abolishing the response to low grades of stretch and partially inhibiting the response to high grades of stretch (69% f 17% p < 0.01). The substance P antagonist [D-Pro2, D-Trp7Yg]substance P partially inhibited ascending contraction induced by high grades of stretch only (40% + 12%, p < 0.02). The vasoactive intestinal peptide antagonist [4-Cl-D-Phe’, Leul’]vasoacfive intestinal peptide inhibited descending relaxation, abolishing the response to low grades of stretch and partially inhibiting the response to high grades of stretch (40% f 4%, p < 0.001). Release of vasoactive intestinal peptide increased significantly by 91% during descending relaxation only, whereas release of both substance P and substance K increased significantly by 107% during ascending contraction only, supporting the participation of vasoactive intestinal peptide motor neurons in descending relaxation and tachykinin motor neurons as well as cholinergic motor neurons in ascending contraction. The components of the human peristaltic reflex and transmitters regulating them were identical to those found in rat and guinea pig intestine. n human and other mammalian species, the myenteric plexus of the enteric nervous system consists of two main populations of neurons that can be distinguished by their content of peptide transmitter. The neurons contain either vasoactive intestinal
I
Virginia
peptide (VIP) and its homologue, peptide histidine methionine (PHM in humans), OI substance P (SP) and its homologue substance K (SK), also known as neurokinin A (1,~). Studies on rat and guinea pig intestine indicate that VIP motor neurons regulate the descending relaxation component of the peristaltic reflex, whereas SP motor neurons regulate in part the ascending contraction component of the reflex (3). Evidence for VIP as the relaxant transmitter in myenteric motor neurons is based on the following: 1. Vasoactive intestinal peptide and its homologues have the unique ability to cause relaxation in circular smooth muscle throughout the digestive tract (4-10). 2. Vasoactive intestinal peptide is released during electrical or reflex activation of myenteric neurons and its release is proportional to the intensity of stimulation and accompanied by a corresponding increase in relaxation (11). induced relaxation is selectively 3. Neurally blocked by specific VIP antiserum and selective VIP antagonists (3,4,6,7,11-17). Evidence for SP [and substance contractile transmitter in myenteric is based on the following:
K (SK]] as a motor neurons
1. Substance P and SK cause direct and cholinergically mediated contraction (18-21). 2. Substance P is released during electrical or reflex activation of myenteric neurons (22-25).
Abbreviations used in this paper: NKA, neurokinin A; PHI, peptide histidine isoleucine; PHM, peptide histidine methionine; SK, substance K; SP, substance P. 0 1989 by the American Gastroenterological Association QQl8-5085/89/$3.59
December
HUMAN PERISTALTIC REFLEX
1989
3. Neurally induced by SP antagonists
contraction is partially blocked (3,20,21,26).
In recent studies on guinea pig and rat (3)) we have used a preparation consisting of an isolated whole segment of small intestine or colon that enables characterization of the components of the peristaltic reflex (ascending contraction and descending relaxation) separately. In the present study, we have modified this preparation to enable recording of the components of the peristaltic reflex from flat sheets of human intestine. The results show that the components of the peristaltic reflex in humans are identical to those in rat and guinea pig. Vasoactive intestinal peptide is released during and is responsible for descending relaxation, whereas acetylcholine is responsible for ascending contraction induced by low grades of stretch and tachykinins are responsible for ascending contraction induced by high grades of stretch.
Materials and Methods Preparation
of Flat
Intestinal
Sheets
A single segment of human jejunum was obtained from each of 22 patients undergoing bypass surgery for morbid obesity; the segments were labeled with a stitch to indicate oral and caudad ends, transported in ice-cold saline, and prepared for the experiments within 30 min of excision. The segments were opened along the mesenteric border and the mucosa was removed by blunt dissection. The segments were on average 6 cm long and 3 cm wide in the unstretched state. The segments were pinned flat on a base of sylgard in an organ bath (Figure 1) containing 10 ml of Krebs’ bicarbonate solution with the following composition (in mM): 118 NaCl, 4.8 KCl, 1.2 KH,PO,, 1.2 MgSO,, 2.5 CaCl,, 25 NaHCO,, 11 glucose, and 0.1% bovine serum albumin. The solution was maintained at 37°C and gassed with 95% O,-5% CO,.
Figure 1 Flat sheet preparation of human intestine for measurement of components of peristaltic reflex. Hook-andpulley assemblies permit radial stretch of circular muscle by application of variable weights (W) to one end and measurement of circular muscle response via a force-displacement transducer at the other end.
Experimental
1415
Procedure
The segments were immobilized so as to enable radial stretch of one end of the segment and recording from the other end using a force displacement transducer (Grass FT03C). Hook-and-pulley assemblies were used to apply radial stretch to and record from the horizontal sheet (27). A set of 4-5 pins secured the midportion
of the segment,
thus separating the stretch site from the recording site. The distance between the stretch site and the recording site was about 4 cm. Radial stretch was applied in the range of 2-14 g at the extreme orad end to elicit descending relaxation only and at the extreme caudad end to elicit ascending contraction only.
Experimental
Design
The jejunal segment was incubated for 1 h, during which the Krebs’ bicarbonate solution was changed four times. The preparation developed basal tone (0.5 g) within 30 min, which remained stable thereafter.
Control stretch-
response curves were then constructed by applying weights in 2-g increments in the range of 2-14 g to the hook-and-pulley assembly. Each increment was applied for a period of 15 s at intervals of 2 min. After 1 h to allow for recovery of the preparation, antagonists were added for 15 min and another stretch-response curve was constructed. The antagonists used included the muscarinic antagonist 1 PM atropine, the nicotinic antagonist 0.1 mM hexamethonium, the VIP/peptide histidine isoleucine (PHI) antagonist 10 PM [&Cl-D-Phe’, Leu17]VIP, and the SP/SK antagonist 10 PM [D-Pro’, D-Trp7*9]SP. The antagonists were used at maximally effective concentrations as determined in previous studies (3,28). Neither the antagonists nor the axonal blocker tetrodotoxin had any effect on basal tone. At the completion of the experiments with antagonists, the bathing medium was changed four times and the preparation retested; in all experiments, the preparation reverted to control responses. For measurement of peptide release in some experiments, the bathing solution was collected during a 15-min basal period and during a subsequent 15-min period in which either ascending contraction or descending relaxation was induced as described by applying increasing stretch for 15 s at intervals of 2 min in the range of 2-12 g. Samples were collected on ice in plastic tubes containing bacitracin (final concentration 20 PM) and aprotinin (1000 U/ml). The samples were divided into l-ml aliquots and stored at - 20°C for subsequent radioimmunoassay. Vasoactive intestinal peptide was measured in duplicate using antiserum AC115 (final concentration 1: 50,000) as described (6,7). The limit of detection of the assay was 5 pg/ml of original sample. The range of concentrations in all samples was 35-220 pg/ml. Substance K was measured in duplicate using SK antiserum 7359N (final concentration 1:2500). The antiserum was specific for SK in that SP cross-reacted only minimally with it (~3%); other peptides such as bombesin and neurotensin did not crossreact. The limit of detection of the assay was 5 pg/ml. The range of concentrations of SK in all samples was 7-105 pg/ml. Substance P was also measured in duplicate using a
1416
GASTROENTEROLOGY Vol. 97, No. 6
GRIDER
ASCENDING
specific SP antiserum, AC95 (final concentration 1:2,500). Substance K, bombesin, neuromedin K, VIP, PHI, dynorphin-13, and somatostatin did not cross-react with this antiserum. The limit of detection of the assay was 10 pg/ml and the IC,, was 130 pg/ml.
Data
DESCENDING
CONTRACTION
RELAXATION
i
Analysis
Descending relaxation and ascending contraction were expressed as percentages of the maximal response obtained using a stretch stimulus of 12 g. Statistical significance was evaluated using Student’s t-test for paired values. Only one experiment was done in each segment. Each experiment consisted of a control measurement in the absence and another in the presence of a specific antagonist; n designates the number of experiments.
Materials [D-Pro’, D-Trp’,‘]SP and [4-Cl-D-Phe’, Leu”lVIP were obtained from Bachem, Torrance, Calif.; VIP antiserum AC115 from Cambridge Research Biochemicals, Valley Stream, N.Y.; SK antiserum 7359N and ?-SK from Peninsula Laboratories, Belmont, Calif.; ‘*‘I-VIP from New England Nuclear, Boston, Mass.; aprotinin from Mobay Chemicals, New York, N.Y.; and bacitracin, atropine, hexamethonium, and bovine serum albumin from Sigma Chemicals, St Louis, MO.
/;
20
I
OL
,
I
I
0
4
CAUDAD
I
I
I
8
STRETCH
1
,
12
(q 1
c
,
,
,
0
4
,
(
, 8
I2
ORAD STRETCH
lq)
Figure 3. Left panel: ascending contraction in response to increasing radial stretch of the caudad end of the human intestinal segment. Right panel: descending relaxation in response to increasing radial stretch of the orad end. Data are expressed as a percentage of maximal response. Values are mean t SE of 7-11 experiments.
ascending contraction response was 0.9 ? 0.1 g (n = 81, and the maximal descending relaxation response was 0.4 ? 0.06 g [n = 11). The pattern was identical to that obtained in whole segments of guinea pig and rat colon (3). Peptide Release During Ascending Contraction and Descending Relaxation
Results Components Reflex
of the Intestinal
Peristaltic
Release of VIP during a 15-min basal period was 96 + 24 pg/g tissue wet weight and increased by
Typical tracings of ascending contraction and descending relaxation are illustrated in Figure 2. Graded radial stretch of the orad end of a flat sheet intestinal preparation elicited increasing descending relaxation only (Figure 3, right). Increasing radial stretch of the caudad end elicited increasing ascending contraction only [Figure 3, left). The maximal
ASCENDING CONTRACTION
DESCENDING RELAXATION
it
T
MD120 -
eo -
b _ h _ - _ - -l_
-h-b-h-
10 P STRETCH
-i
H
;
-+
q-
60 IIJi %3 w J= LlJO aL
4O20 -
/o.59
Figure 2. Upper panel: typical ascending contraction response elicited by radial stretch of the caudad end of the human intestinal segment. Lower panel: typical descending relaxation response elicited by radial stretch of the caudad end of the human intestinal segment. Numbers in the center indicate the grams of stretch applied via the hook-and-pulley assembly. The horizontal bars indicate the period of stretch (15 s).For the remainder of the 2-min period before application of the next stimulus, the preparation reverted to basal level.
-60
t
1
Figure 4. Release of VIP, SK, and SP during ascending contraction and descending relaxation. Data are expressed as a percentage change from basal release. Values are mean 2 SE of 3-6 experiments.
December 1989
HUMAN PERISTALTIC REFLEX
ASCENDING CONTRACTION
ASCENDING CONTRACTION
DESCENDING RELAXATION
VIP
4
0 CAUDAD
Figure
6 STRETCH
DESCENDING RELAXATION
ANTAGONIST
12
VIP ANTAGONIST
((1)
ORAD
STRO%H
(g)
5. Left panel: ascending contraction in response to increasing radial stretch of the caudad end of human intestinal segment in the absence (closed circles] and presence (open circles) of atropine (1 PM). Right panel: descending relaxation in response to increasing radial stretch of the orad end in the absence [closed circles] and presence (open circles) of atropine (1 PM). Data are expressed as a percentage of maximal response. Values are mean 2 SE of 4 experiments.
91% k 28% (p < 0.02;n = 6)during a 15-min period in which descending relaxation was induced by progressive radial stretch of the extreme orad end (Figure 4). In contrast, release of VIP decreased by 31% + 23% (NS;n = 3) during a 15-min period in which ascending contraction was induced by progressive radial stretch of the extreme caudad end (Figure 4). The pattern of release of SK was the reverse of that for VIP (Figure 4). Release of SK during a 15-min basal period was 59 ? 8 pg/g tissue wet weight and increased by 107% + 32% (p < 0.05) during a 15-min period in which ascending contraction was induced by progressive radial stretch of the extreme caudad 100
1417
CONTROL
HUMAN JEJUNUM
SP ANTAGONIST
ATROPINE
OL
0
1
1
4
CAUDAD Figure
7.
*
j
8
STRETCH
j
c
12 (g)
1
0
4 ORAD
6 STRETCH
12 (g 1
Effect of VIP antagonist {B-Cl-D-Phe’, Leu171VIP (10 PM)} on ascending contrktion (left panel) kd descending relaxation (right panel]. Data are expressed as a percentage of maximal response. Values are mean 2 SE of 3-5 experiments.
end. In contrast, release of SK decreased by 12% + 9% (NS) during a 15min period in which descending relaxation was induced by progressive radial stretch of the extreme orad end. The pattern of release of SP was identical to that of SK. During the 15min period in which ascending contraction was induced, release of SP increased by 108% 2 46% above basal level (p < 0.05), whereas during the 15-min period in which descending relaxation was induced, SP release was decreased by 14% + 17% (NS] Efiects of Nicotinic and Muscarinic Antagonists on the Components of the Peristaltic Reflex Tetrodotoxin (1 PM) and hexamethonium (100 PM) abolished both ascending contraction and descending relaxation at all grades of stretch. Atropine (1 PM) had no effect on descending relaxation but inhibited ascending contraction at all grades of stretch (Figure 5). The ascending contraction response to 2- and 4-g stimuli was abolished, whereas the response to 10 g was inhibited by 69% 2 17% (p < 0.01; n = 4).
ANTA%NlST ATR;PlNE 0
4 CAUDAD
Figure
6.
8 STRETCH
12 (g)
Effect of atropine (1 PM), SP antagonist {[D-Pro’, DTrp7,‘]SP (10 j&f)}, and a combination of atropine and SP antagonist on ascending contraction induced by various grades of stretch. Data are expressed as a percentage of maximal response. Values are mean 2 SE of 4 experiments.
Effect of Tachykinin Antagonist on the Components of the Peristaltic Reflex The tachykinin antagonist [D-Pro’, D-Trp’,‘]SP (10 PM), like atropine, had no effect on descending relaxation, but inhibited ascending contraction (Figure 6).However, unlike atropine, it had little or no effect on ascending contraction induced by the lowest grades of stretch (2 and 4 g), but inhibited
1418
GASTROENTEROLOGY
GRIDER
ascending contraction induced by the higher grades of stretch; the response to 12 g of stretch was inhibited by 40% -+12% (p < 0.02;n = 4) (Figure 6).The combination of atropine and tachykinin antagonist abolished the response to low grades of stretch and inhibited the response to higher grades of stretch by 84% + 6% (p < 0.01;n = 4).After removal of the antagonists from the bathing medium by repeated washing, the responses returned to control levels. Effect of Vasoactive Intestinal Peptide Antagonist on the Components of the Peristaltic Reflex The VIP antagonist [4Cl-D-Phe”, Leu’7]VIP (10 PM) inhibited descending relaxation at all grades of stretch (Figure 7). The response to the lowest grades of stretch (2 and 4 g) was abolished and the response to the highest grade of stretch (12 g) was inhibited by 40% + 4% (p < 0.001;n = 5). In contrast, the VIP antagonist augmented ascending contraction at all grades of stretch. The response to the lowest grade of stretch (2g)was increased by 79% + 4% (p < 0.01, n = 3) and the response to the highest grade of stretch was increased by 19% t 1% (p < 0.01;n = 3).
Discussion This study identifies for the first time the peptide transmitters involved in the regulation of the intestinal peristaltic reflex in humans. For this purpose, small segments of human intestine obtained at surgery were used as flat sheets to examine separately the components of the peristaltic reflex. The components of the reflex (ascending contraction and descending relaxation) were shown to be identical to those in the colon of the rat and guinea pig (3). The identity of the peristaltic reflex in humans and animals extends to the following. First, in both human and animal intestine (3,27), the pathways regulating ascending contraction and descending relaxation involve the participation of cholinergic interneurons, as both components are abolished by nicotinic antagonists. Second, cholinergic motor neurons are involved in the regulation of ascending contraction only (Figure 5). Ascending contraction induced by low grades of stretch is abolished by atropine, implying that at this intensity of stimulation cholinergic motor neurons only are involved. Third, tachykinin neurons are involved in the regulation of ascending contraction induced by high grades of stretch; at this intensity of stimulation, both tachykinin and cholinergic neurons are involved because ascending contraction induced by high grades of stretch is sensitive to both atropine
Vol. 97, No. 6
and SP antagonists (Figure 6).In human as in animal intestine, ascending contraction only is accompanied by SK and SP release. The release of both SP and SK implies that the tachykinin precursor in these motor neurons is /?-preprotachykinin which contains the sequences of both peptides. Fourth, VIP motor neurons regulate descending relaxation. In human and animal intestine (3), descending relaxation only is accompanied by release of VIP. The VIP release is coupled to descending relaxation because selective VIP antagonists inhibit descending relaxation (28).In animal intestine, in addition, descending relaxation is selectively inhibited by VIP-specific antiserum in a concentrationdependent manner (3). It is noteworthy that descending relaxation induced by low grades of stretch in human and animal intestine is abolished by the VIP antagonist, whereas descending relaxation induced by high grades of stretch is partially inhibited consistent with competitive antagonism. As shown previously (28),the antagonist used in the present study inhibited relaxation induced by both VIP and PHI. The extent of the contribution of PHM (or PHI) is not known, but is likely to be minor as PHUPHM, though derived from the same precursor as VIP, is more extensively processed in tissues and is less potent (5-10 times) than VIP as a relaxant agent (29,30). The augmentation of ascending contraction by VIP antagonist, which is also observed with VIP antiserum in animal studies, implies that background VIP influences the contractile state and response of circular smooth muscle. Recent studies support the notion that the inhibitory activity of VIP neurons masks myogenic phasic activity, which can be restored by addition of VIP antiserum or VIP antagonists (31). Thus, the influence of cholinergic and tachykinin neurons in the regulation of ascending contraction should be viewed in relation to the background influence of VIP neurons.
References Wattchow DA, Furness JB, Costa M. Distribution and coexistence of peptides in nerve fibers of the external muscle of the human gastrointestinal tract. Gastroenterology 1988;95:3241. Llewellyn-Smith IJ, Furness JB, Costa M, Gibbins IL. Quantitative ultrastructural analysis of enkephalin-, substance Pand VIP-immunoreactive nerve fibers in the circular muscle of the guinea pig small intestine. J Comp Neurol 1988; 272:139-48. Grider JR, Makhlouf GM. Colonic peristaltic reflex: identification of vasoactive intestinal peptide as mediator of descending relaxation. Am J Physiol 1986;251:G40-5. Biancani P, Walsh JH, Behar J. Vasoactive intestinal peptide: a neurotransmitter for relaxation of the rabbit internal anal sphincter. Gastroenterology 1985;89:867-74.
December 1989
5. Bitar KN, Makhlouf GM. Relaxation of isolated gastric smooth muscle cells by vasoactive intestinal peptide. Science (Wash DC) 1982:261:531-3. 6. Grider JR, Cable MB, Said SI, Makhlouf GM. Vasoactive intestinal peptide (VIP] as neural mediator of gastric relaxation. Am J Physiol 1985;248:G73-8. 7. Grider JR, Cable MB, Bitar KN, Said SI, Makhlouf GM. Vasoactive intestinal peptide. Relaxant neurotransmitter in tenia coli of the guinea pig. Gastroenterology 1985;89:36-42. 8. Makhlouf GM. Enteric neuropeptides: role in neuromuscular activity of the gut. Trends Pharmacol Sci 1985;6:214-8. 9. Larsson LT, Malmfors G, Wahlestedt C, Leander S, Hakanson R. Hirschprung’s disease: a comparison of the nervous control of ganglionic and aganglionic smooth muscle in vitro. J Pediatr Surg 1987;22:431-5. 10. Furness JB, Costa M. VIP and enteric inhibitory nerves. In: Said S, ed. Vasoactive intestinal peptide. New York: Raven, 1982:391-406. 11. Grider JR, Makhlouf GM. Prejunctional inhibition of vasoactive intestinal peptide release. Am J Physiol 1987;253:G7-12. 12. Angel F, Go VLW, Schmalz PF, Szurszewski JH. Vasoactive intestinal polypeptide. A putative neurotransmitter in the canine gastric muscularis mucosae. J Physiol Lond 1983; 341:641-5. 13. Biancani P, Walsh JH, Behar J. Vasbactive intestinal polypeptide. A neurotransmitter for lower esophageal sphincter relaxation. J Clin Invest 1984;73:963-7. 14. D’Amato M, DeBeurme FA, Lefebvre A. Comparison of the effect of vasoactive intestinal polypeptide and non-adrenergic, non-cholinergic neurone stimulation in the cat gastric fundus. Eur J Pharmacol 1988;152:71-82. 15. Goyal RK, Rattan S, Said SI. VIP as a possible neurotransmitter of non-cholinergic, non-adrenergic inhibitory neurons. Nature 1985;288:378-80. 16. Nurko S, Rattan S. Role of vasoactive intestinal polypeptide in the internal anal sphincter relaxation of the opossum. J Clin Invest 1988;81:1146-53. 17. Wiley JW, O’Dorisio TM, Owyang C. Vasoactive intestinal polypeptide mediates cholecystokinin-induced relaxation of the sphincter of Oddi. J Clin Invest 1988:81:1920-l. 18 Souquet JC, Bitar KN, Grider JR, Makhlouf GM. Receptors for substance P on isolated intestinal smooth muscle cells of the guinea pig. Am J Physiol 1987;253:G666-72. 19. Fosbraey P, Featherstone RL, Morton IKM. Comparison of potency of substance P and related peptides on 13H]acetylcholine release and contractile actions in the guinea pig ileum. Naunyn Schmiedebergs Arch Pharmacol 1984;326: 111-5. 20 Kilbinger H, Staub P, Erlhof I, Holzer P. Antagonist discimi-
HUMAN PERISTALTIC REFLEX
1419
nation between subtypes of tachykinin receptors in the guinea pig. Naunyn Schmiedebergs Arch Pharmacol 1986; 334:181-7. 21. Costa M, Furness JB, Pullin CO, Bornstein J. Substance P enteric neurons mediate non-cholinergic transmission to the circular muscle of the guinea pig intestine. Naunyn Schmiedebergs Arch Pharmacol 1985;328:446-63. 22. Donnerer J, Barth0 L, Holzer P, Lembeck F. Intestinal peristalsis associated with the release of immunoreactive substance P. Neuroscience 1984;11:913-18. 23. Barson SA, Jaffe BM, Gintzler AR. Release of substance P from the enteric nervous system: direct quantitation and characterization. J Pharmacol Expt Ther 1983;227:365-8. 24. Grider JR, Makhlouf GM. Regulation of the ascending contraction component of the peristaltic reflex by myenteric tachykinin neurons (abstr). Gastroenterology 1988;94:A157. release of 25. Holzer P. Characterization of the stimulus-induced immunoreactive substance P from the myenteric plexus of the guinea pig small intestine. Brain Res 1984;297:127-36. 26. Barth0 L, Holzer P, Donnerer J, Lembeck F. Effects of substance P, cholecystokinin octapeptide, bombesin, neurotensin on the peristaltic reflex of the guinea pig ileum in the absence and presence of atropine. Naunyn Schmiedebergs Arch Pharmacol 1982;321:321-8. 27. Costa M, Furness JB. The peristaltic reflex: an analysis of the nerve pathways and their pharmacology. Naunyn Schmiedebergs Arch Pharmacol 1976;294:47-60. 28. Grider JR, Rivier J, Makhlouf GM. Evidence for VIP as transmitter of inhibitory motor neurons: blockade of neurally mediated relaxation by VIP antagonists (abstr). Gastroenterology 1987;92:1415. 29. Fahrenkrug J, Bek T, Lundberg JM, Hokfelt T. VIP and PHI in cat neurons: colocalization but variable tissue content possibly due to differential processing. Regul Pept 1985;12:21-34. 30. Itoh N, Obata K, Yanihara N, Okamoto H. Human preprovasoactive intestinal peptide contains a novel PHI-27-like peptide, PHM-27. Nature 1983;304:547-9. 31 Grider JR, Makhlouf GM. Suppression of inhibitory neural input to colonic circular muscle by opioid peptides. J Pharmacol Exp Ther 1987;243:205-10.
Received January 6, 1989. Accepted May 31,1989. Address requests for reprints to: J. R. Grider, Ph.D., Box 711, MCV Station, Medical College of Virginia, Richmond, Virginia 23298-711. This work was supported by grant DK34153 from the National Institute of Diabetes, and Digestive and Kidney Diseases.