Abstract
The neurotransmitters mediating relaxation of lower esophageal sphincter (LES) were studied using circular LES strips from adult pigs in organ baths. LES relaxation by sodium nitroprusside (1 nM–3 μM), vasoactive intestinal peptide (VIP) and pituitary adenylate cyclase-activating peptide (PACAP; 1 nM–1 μM), ATP (10 μM–30 mM), and tricarbonyldichlororuthenum dimer (1 μM–1 mM) was unaffected by tetrodotoxin (1 μM) or l-NG-nitroarginine methyl ester (l-NAME; 100 μM). Calcitonin gene-related peptide (CGRP; 1 nM–1 μM) did not affect LES tone. ATP relaxation was blocked by 1 μM apamin and the P2Y1 antagonist MRS 2179 (N6-methyl 2′-deoxyadenosine 3′,5′-bisphosphate; 10 μM). Apamin inhibited PACAP relaxation. VIP and PACAP relaxation was blocked by 10 U/ml α-chymotrypsin. l-NAME (–62.52 ± 13.13%) and 1H-[1,2,4]oxadiazole-[4,3-α]quinoxalin-1-one (ODQ; 10 μM, –67.67 ± 6.80%) similarly inhibited electrical LES relaxation, and apamin blocked non-nitrergic relaxation. Nicotine relaxation (100 μM) was inhibited by l-NAME (–60.37 ± 10.8%) and ODQ (–41.90 ± 7.89%), and apamin also blocked non-nitrergic relaxation. Non-nitrergic and apamin-sensitive LES relaxation by electrical stimulation or nicotine was strongly inhibited by MRS 2179, slightly inhibited by α-chymotrypsin and the P2X1,2,3 receptor antagonist NF 279 (8,8¢-[carbonylbis(imino-4,1-phenylenecarbonylimino-4,1-phenylenecarbonylimino)]bis-1,3,5-naphthalenetrisulfonic acid hexasodium salt; 10 μM), and unaffected by tin protoporphyrin IX (100 μM). Porcine LES relaxation after stimulation of intrinsic inhibitory motor neurons is mediated by two main neuromuscular pathways: nitric oxide through guanylate cyclase signaling and apamin-insensitive mechanisms and by non-nitrergic apamin-sensitive neurotransmission mainly mediated by ATP, ADP, or a related purine acting on P2Y1 receptors and a minor contribution of purinergic P2X1,2,3 receptors and PACAP. Nitrergic and purinergic co-transmitters show parallel effects of similar magnitude without major interplay. Our study shows no role for CGRP and only a minor one for VIP and carbon monoxide in porcine LES relaxation.
The lower esophageal sphincter (LES) acts as a barrier at the gastroesophageal junction (GEJ). Basal tone of LES is primarily myogenic in origin and is modulated by a combination of hormonal factors and neurogenic mechanisms that involve local enteric motor neurons and extrinsic nerves and are not yet fully understood. Nonadrenergic, noncholinergic enteric motor neurons are the final step in the inhibitory vagal pathway to LES, allowing swallowing-induced and transient LES relaxation that causes physiologic gastroesophageal reflux and belching (Chang et al., 2003). Vasoactive intestinal peptide (VIP), nitric oxide, ATP, pituitary adenylate cyclase-activating peptide (PACAP), carbon monoxide, and calcitonin gene-related peptide (CGRP) have been proposed as putative neurotransmitters for these inhibitory enteric motor neurons in LES on the basis of morphological and physiological studies. Colocalization of these neurotransmitters and/or their synthesizing enzymes on inhibitory enteric motor neurons of the upper gastrointestinal tract including LES has been described by immunohistological and other morphological studies (Ny et al., 1994, 1995a; Uc et al., 1997; Werkstrom et al., 1997). Physiological and pharmacological studies have demonstrated the direct effects or characterized the actions of these neurotransmitters after stimulation of inhibitory LES motor neurons (Ny et al., 1995b, 1997; Uc et al., 1997; Yuan et al., 1998; Imaeda and Cunnane, 2003). However, evidence suggests that nitric oxide from neural sources is the major contributor to LES relaxation (Murray et al., 1991; Tottrup et al., 1991; Gonzalez et al., 2004), and the relative physiological contribution of other neurotransmitters remains unclear. On the other hand, several mechanisms of interaction between these neurotransmitters have been proposed: release “in parallel” and independent actions on specific postjunctional sites (cotransmission) (Burnstock, 2004), effects coupled “in series” with nitric oxide-mediated release of other neurotransmitters (Grider et al., 1992), or pre- and/or postjunctional modulation of nitric oxide synthesis by other neurotransmitters (Mashimo et al., 1996; Ergun and Ogulener, 2001). The underlying mechanisms and physiological relevance of these interactions on LES have not been discovered. Most of these studies focused on the mechanisms of LES relaxation during simultaneous and direct electrical stimulation of excitatory and inhibitory motor neurons, and little data are available on the effects of these inhibitory neurotransmitters following other more specific stimuli for the inhibitory neurons. In a recent in vitro study on human LES, we found that inhibitory enteric motor neurons are efficiently stimulated both by electrical field stimulation (EFS) and nicotinic acetylcholine receptors (AChRs) located in somatodendritic regions and nerve terminals, whereas esophageal excitatory motor neurons are also efficiently stimulated by EFS, but their stimulation through nicotinic AChRs is difficult and causes a weak response (Gonzalez et al., 2004).
The porcine gastrointestinal tract possesses anatomic and pathological similarities to that of humans and similar organization of the enteric nervous system, differing from small laboratory animals, and has been used as a homologous animal model for the development of new pharmacologic strategies to treat human neurogastrointestinal disorders (Pasricha et al., 1993; Brown and Timmermans, 2004). The size, histology (smooth muscle cells), and neurochemical code of porcine LES motor neurons are similar to that of humans (Aggestrup et al., 1986; Pasricha et al., 1993). The aim of the present study was to characterize the neuromyogenic mechanisms and neurotransmitters controlling tone and relaxation of porcine LES after stimulation of inhibitory enteric motor neurons by EFS and through nicotinic AChRs.
Materials and Methods
Preparations
Specimens including part of the gastric fundus, the GEJ, and the esophageal body were obtained from 53 adult pigs (age, 6 months; weight, 75–80 kg). Animals were stunned and killed by exsanguination in a slaughterhouse in compliance with specific national laws following the guidelines of the European Union. Specimens were immediately collected, placed in carbogenated Krebs' solution at 4°C, and transported to the laboratory within 1 h. The GEJ was opened along the greater curvature, the mucosa and submucosa were resected at the squamocolumnar union, and clasp and sling fibers composing LES were identified (Preiksaitis and Diamant, 1997). Full-thickness preparations including the circular and longitudinal muscle layers as well as the myenteric plexus were obtained by cutting 3-mm-wide strips parallel to circular muscle fibers from the clasp region of the LES of each specimen.
Procedures
Studies started within 18 h of sacrifice. Strips measuring 10 mm in length were placed in 15-ml organ baths containing Krebs' solution constantly bubbling with 5% CO2 in O2. Changes in tension of the strips were measured using isometric force transducers, recorded on a chart recorder (model 03 Force Transducer and model 7 Series Polygraph, respectively; Grass Instruments Co., Quincy, MA), and digitized (AcqKnowledge MP100; BIOPAC Systems, Inc., Goleta, CA). After an equilibration period of 30 min, strips were stretched up to 150% of their initial length and positioned between two parallel platinum wire electrodes 10 mm apart. Most strips taken from the GEJ progressively increased their tension during the following 1 to 2 h. This increase in tension was defined as the active resting tone (Tottrup et al., 1990). EFS was applied by means of an electrical stimulator (model S88; Grass Instruments Co.) and a power booster (Stimu-Splitter II; Med-Lab Instruments, Loveland, CO) to obtain six identical and undistorted signals. Only the strips that developed active tension during the equilibration period and relaxed with EFS and/or nicotine were considered as pertaining to the LES and included in the study.
Solutions and Drugs
The Krebs' solution used in these experiments contained 138.5 mM Na+, 4.6 mM K+, 2.5 mM Ca2+, 1.2 mM Mg+, 125 mM Cl–, 21.9 mM HCO –3, 1.2 mM H2PO –4, 1.2 mM SO –4, and 11.5 mM glucose. Sodium nitroprusside, VIP, ATP, CGRP, apamin, nicotine, hexamethonium, l-NAME, suramin, VIP 6–28, 1H-[1,2,4]oxadiazole-[4,3-α]quinoxalin-1-one (ODQ), α-chymotrypsin, and tricarbonyl dichlororuthenum dimer (CORM-1) were obtained from Sigma-Aldrich Co. (Madrid, Spain). Tetrodotoxin was purchased from Latoxan (Valence, France), PACAP 28 was purchased from Peptides Institute (Osaka, Japan), and the competitive antagonist for P2Y1 receptors MRS 2179, the antagonist for the P2X1,2,3 receptors NF 279, the selective heme oxygenase inhibitor tin protoporphyrin IX (SnPP-IX), and the P2Y1 agonist adenosine 5′-O-2-thiodiphosphate (ADPβS) were purchased from Tocris Cookson Ltd. (Bristol, UK). All drugs were dissolved in distilled water, with the exception of ODQ, which was dissolved in ethanol <5% (v/v), and CORM-1 and SnPP-IX, which were freshly prepared before each experiment and dissolved in DMSO. Final concentration of DMSO in the organ bath did not exceed 0.1% (v/v). Experiments using SnPP-IX were conduced in the dark. In pilot studies, we found that at these concentrations, ethanol and DMSO did not significantly alter resting tone or EFS-induced responses (data not shown).
Experimental Design
To ascertain the physiological relevance of each putative neurotransmitter on LES physiology, our experimental design included the effect of neurotransmitters on LES tone, the selection and characterization of specific antagonists for each neurotransmitter, and the evaluation of their possible interactions, particularly the possible induction of nitric oxide synthesis by any other neurotransmitter; the study of neuromyogenic mechanisms controlling active LES tone; and the characterization of the relaxatory responses after stimulation of LES enteric inhibitory motor neurons by EFS or through nicotinic AChRs, the effects of the neurotransmitters released, and the assessment of the possible nitric oxide-mediated release of any other putative neurotransmitters.
The Effect of Neurotransmitters on LES Tone and the Selection of Specific Antagonists. Concentration-related curves of the effects of sodium nitroprusside, VIP, PACAP, ATP, and CGRP were obtained by exposing LES strips to single doses of agonists for up to 3 min. After washing the strips with 45 ml of fresh buffer, there was a 30-min period before the next exposure. Concentration-response curves for CORM-1 were obtained in a cumulative fashion. Concentrations of neurotransmitters were selected according to previous studies, most on LES (Yamato et al., 1992; Ny et al., 1995b; Uc et al., 1997, 1999; Rattan et al., 2004). We used 1 nM to 3 μM for sodium nitroprusside, 1 nM to 1 μM for VIP, PACAP, and CGRP; 10 μM to 30 mM for ATP; and 1 μM to 1 mM for CORM-1. Reported concentrations are final bath concentrations. Responses to 10 μM to 1 mM ADPβS were also assessed to further characterize the purinergic receptors involved in LES relaxation (De Man et al., 1991; Yoshioka and Nakata, 2004). Submaximal LES responses to sodium nitroprusside, ATP, VIP, PACAP, and CORM-1 and ADPβS were studied in the presence of the neurotoxin tetrodotoxin (1 μM) and the following substances: ODQ (10 μM), a specific inhibitor of the soluble guanylyl cyclase that has been widely used as a pharmacological tool to characterize the NO-mediated component of EFS-induced LES relaxation (Shahin et al., 2000); apamin (1 μM), blocker of small conductance Ca2+-activated K+ channels; MRS 2179 (10 μM), competitive antagonist for P2Y1 receptors (Boyer et al., 1998; De Man et al., 2003) and NF 279 (10 μM) for P2X1,2,3 receptors (Damer et al., 1998; De Man et al., 2003); the peptidase α-chymotrypsin (10 U/ml) that cleaves VIP and PACAP at the level of tyrosine residues (Mule and Serio, 2003) and the VIP antagonist, VIP 6–28 (100 nM); and also following the nitric oxide synthase inhibitor l-NAME (100 μM) to explore the possible involvement of nitric oxide in the response of other neurotransmitters. Antagonists were added to the baths 30 min before addition of neurotransmitters with the exception of α-chymotrypsin, which was added 10 min before. In preliminary experiments, we found that, at the dose and time of exposition used, α-chymotrypsin did not affect responses to repetitive doses of VIP or PACAP (data not shown).
Control of Active LES Resting Tone. The neurogenic versus myogenic contribution to active LES resting tone was assessed by measuring changes in tone after exposure to the neurotoxin tetrodotoxin 1 μM and to a Ca2+-free medium for 30 min. The influence of tonic activity of inhibitory motor neurons upon LES resting tone was studied by measuring maximal changes in tone after exposure to 100 μM l-NAME, 10 μM ODQ, 1 μM apamin, and 10 μM MRS 2179 for 30 min or α-chymotrypsin 10 U/ml for 10 min. The influence of cholinergic motor neurons upon LES tone was assessed by atropine 1 μM for 30 min.
Stimulation of Enteric Motor Neurons by EFS, the Effects of the Neurotransmitters Released, and Characterization of Nitric Oxide-Mediated Effects. Transmural EFS (pulses of 0.4-Ms duration, frequency, 0.3–20 Hz) were applied to LES preparations in 5-s trains at 26 V (Gonzalez et al., 2004). The amplitude of EFS-induced responses did not decay during control experiments lasting up to 4 h (data not shown). The neural origin of EFS-induced responses was assessed by 1 μM tetrodotoxin and 100 μM hexamethonium. First, two experiments were conducted during direct stimulation of inhibitory motor neurons by EFS and sequential addition of antagonists to assess the nature and effects of the neurotransmitters released. In the first experiment (Experiment E-1), the nitrergic and cholinergic components of EFS responses were sequentially blocked by l-NAME and 1 μM atropine, and apamin was assessed on non-nitergic relaxation. Results were compared with experiment E-2 during blockade by 10 μM ODQ of the target (guanylate cyclase) of the synthesized and released nitric oxide, allowing the effect of any neurotransmitter coupled in series with nitric oxide, or released by nitric oxide. In experiments E-3 to E-6, the effect of MRS 2179 (experiment E-3), NF 279 (experiment E-4), α-chymotrypsin (experiment E-5), and 100 μM SnPP-IX (experiment E-6) on l-NAME-resistant relaxation was assessed. Drugs were added to the baths 30 min before EFS with the exception of α-chymotrypsin, which was added 10 min before.
Stimulation of Enteric Motor Neurons with Nicotine, Identification of the Neurotransmitters Released, and Characterization of Nitric Oxide-Mediated Effects. A concentration-related curve of the effect of nicotine (1–300 μM) on LES strips was obtained by exposing the strips for 2 min to single doses of the drug to assess the LES relaxation induced by stimulation of enteric motor neurons through nicotinic AChRs. Strips were washed with 45 ml of fresh buffer and left for 30 min before exposure. Repeated additions of nicotine (100 μM) did not desensitize nicotinic receptors (data not shown). The specificity of the effects of nicotine was assessed by the ganglionic blocker, 100 μM hexamethonium, and the site of effect by 1 μM tetrodotoxin (Galligan, 1999; Gonzalez et al., 2004). Nicotine dose, which caused a submaximal LES relaxation, was first selected for the studies with antagonists to characterize the neurotransmitters released by stimulation of inhibitory motor neurons through nicotinic AChRs. In the first experiment (N-1), the nitrergic component of nicotine response was assessed by l-NAME, and apamin was assessed on non-nitergic relaxation. Results were compared with experiment N-2 during blockade by 10 μM ODQ of the target (guanylate cyclase) of the synthesized and released nitric oxide, allowing the effect of any neurotransmitter coupled in series with nitric oxide or released by nitric oxide. In experiments N-3 to N-6, the relative effect of MRS 2179 (experiment N-3), NF 279 (experiment N-4), α-chymotrypsin (experiment N-5), and SnPP-IX (experiment N-6) on l-NAME-resistant relaxation was assessed. Drugs were added to the baths 30 min before the stimulation of inhibitory motor neurons with nicotine, with the exception of α-chymotrypsin, which was added 10 min before. Experiments were conduced at basal conditions and not in nonadrenergic, noncholinergic conditions to compare neural stimulation by EFS or with that of nicotinic AChRs (Gonzalez et al., 2004).
Data Analysis
The effects of EFS and pharmacological agents were determined in terms of changes in tone. Relaxation was expressed in grams and/or in the percentage of active LES resting tone (Gonzalez et al., 2004). The dose-response curve was computer-fitted using nonlinear regression and the maximal response elicited by the agonist, and the EC50 values were calculated (GraphPad Prism, version 2.1; GraphPad Software Inc., San Diego, CA). Contraction was expressed in grams and/or in percentage of active LES resting tone. Data are expressed as mean ± mean S.E. Student's t test was selected for comparisons, using the paired mode when appropriate, and the effect of pharmacological agents on frequency-response curves was performed using two-way repeated measure analysis of variance. When the Student's t test was significant, the Bonferroni test was carried out to determine the frequencies of statistically different responses. p < 0.05 was considered statistically significant.
Results
LES strips developed an active resting tone of 4.90 ± 0.30g (n = 34). Sodium nitroprusside, VIP, PACAP, ATP, ADPβS, and CORM-1 relaxed LES strips; in contrast, CGRP had no effect at any dose tested (Figs. 1 and 2). The nitric oxide donor sodium nitroprusside, the carbon monoxide donor CORM-1, and VIP (Fig. 1) induced a monophasic LES relaxation. PACAP and ADPβS induced a biphasic response with an initial fast followed by a slow and sustained relaxation (Fig. 1). ATP (1 mM) evoked a triphasic mechanical response including an initial fast relaxation, which was followed by a fast contraction and a slow sustained relaxation (Fig. 1). Relaxation induced by sodium nitroprusside, CORM-1, and VIP was slower than initial fast relaxation induced by PACAP and ATP (Table 1). The nitric oxide donor, VIP, PACAP, CORM-1, ATP, and ADPβS relaxed LES strips in a concentration-dependent manner (Fig. 2). Table 1 summarizes the dynamics of the effects of proposed neurotransmitters on LES. PACAP and nitric oxide induced a complete LES relaxation at micromolar concentrations and showed EC50 in the 10 nM range. ATP induced a complete relaxation at millimolar concentrations. VIP relaxed the active tone of strips by only 47 ± 12.2% at micromolar concentrations and CORM-1 by 30.94 ± 6.7% at millimolar concentrations.
The doses of neurotransmitters causing submaximal LES relaxation selected for studies with antagonists were sodium nitroprusside (1 μM), VIP (100 nM), CORM-1 (500 μM), ATP (1 mM), ADPβS (10 μM), and PACAP (100 nM), the effects of antagonists having been evaluated in individual experiments using three to seven specimens. Inhibition of nitric oxide synthesis by l-NAME did not significantly modify any of the responses induced by sodium nitroprusside (+0.50 ± 4.02%, N.S.), VIP (–8.56 ± 2.91%, N.S.), ATP (fast relaxation, +0.30 ± 4.86%; fast contraction, –13.24 ± 4.47%; slow relaxation, +4.01 ± 11.27%, N.S.), or PACAP (fast relaxation, –11.64 ± 10.92%; slow relaxation, –13.40 ± 5.28%, N.S.). Nitric oxide-induced relaxation by sodium nitroprusside was specifically blocked by ODQ (–98.35 ± 1.15%, p < 0.05) and unaffected by the other antagonists. Fast relaxation induced by ATP was unaffected by ODQ (+4.00 ± 3.13%, N.S.), l-NAME (+0.30 ± 4.86%, N.S.), or α-chymotrypsin (+0.26 ± 8.04%, N.S.). In contrast, apamin fully blocked (–98.84 ± 1.16%, p < 0.01) the initial fast relaxation and switched the LES response to ATP into a biphasic response with an initial contraction of 1.68 ± 0.78 g and a slow sustained relaxation. The ATP-induced contraction was not affected either by tetrodotoxin (–0.44 ± 16.18%, N.S.), or ODQ (–12.18 ± 6.05%, N.S.). MRS 2179 (10 μM) strongly inhibited fast ATP-induced relaxation (–76.3 ± 4.88%, p < 0.05) and fast contraction (–80.38 ± 0.75%, p < 0.05). Apamin inhibited fast component (–76.78 ± 6.06%, p < 0.05) and had no effect on the slow component (–31.75 ± 9.49, N.S.) of PACAP-induced LES relaxation. Unexpectedly, CORM-1 relaxation was unaffected by either ODQ or apamin. VIP 6–28 (100 nM) did not significantly affect VIP relaxation (+1.78 ± 7.02%, N.S.), and α-chymotrypsin fully blocked VIP and PACAP-induced relaxation (–100%, p < 0.05). Tetrodotoxin did not significantly affect the relaxation induced by sodium nitroprusside (–3.04 ± 2.17%, N.S.), VIP (–5.62 ± 5.49%, N.S.), ATP (fast relaxation, +27.31 ± 9.21%; slow relaxation, +21.30 ± 9.57%, N.S.), ADPβS (10 μM) (fast relaxation, +6.84 ± 9.81%; slow relaxation, +15.32 ± 14.45%, N.S.), and PACAP (fast relaxation, –20.26 ± 16.89%, N.S.; slow relaxation, –30.57 ± 8.93%, N.S.).
Control of Active LES Resting Tone.Figure 3 illustrates the myogenic and neurogenic factors contributing to LES tone. Exposure of LES strips to a Krebs' Ca2+-free buffer fully abolished active resting tone by 96.40 ± 5.82% (n = 9). Simultaneous blockade of intrinsic excitatory and inhibitory neural inputs by tetrodotoxin did not significantly affect active LES tone (–6.93 ± 8.84% n = 7, N.S.). In contrast, atropine significantly reduced active LES resting tone by –24.3 ± 3.1% (n = 17) (p < 0.05), and l-NAME, ODQ and apamin significantly enhanced active LES resting tone by 46.45 ± 3.31% (n = 17), 28.22 ± 2.83% (n = 19), and 20.3 ± 2.1% (n = 12), respectively (p < 0.05); MRS 2179 enhanced LES tone by only 2.87 ± 2.16% (n = 6, N.S.). The enhancement caused by l-NAME on LES tone was higher than that caused by ODQ or apamin (p < 0.001). α-Chymotrypsin exerted a variable effect on LES because active tone was either enhanced by 41 ± 47% (n = 8) or reduced by 36.81 ± 47% (n = 7) after 10 min of exposure to the drug. These experiments show that porcine LES tone is mainly myogenic, depends on extracellular Ca2+, and is modulated by tonic input from cholinergic motor neurons and by continuous influence of inhibitory motor neurons.
Inhibitory Neurotransmitters in EFS-Induced Relaxation. LES strips responded to EFS with a sharp relaxation during electrical stimulus (“on” relaxation) followed by a phasic contraction at the end of the stimulus (“off” contraction) (Fig. 4). The amplitude of both responses was frequency-dependent, and maximal relaxation (4.21 ± 0.25 g or 90.4 ± 3.1% of resting tone) was reached at 3 Hz (n = 34). EFS-induced relaxation was fully blocked by tetrodotoxin at all frequencies tested (–100%, n = 5, p < 0.001) and unaffected by hexamethonium (+6.78 ± 14.65% at 3 Hz, n = 3, N.S.), showing the origin of these responses in the activation of enteric motor neurons. In experiment E-1 (Fig. 4), l-NAME significantly inhibited EFS relaxation with significant effects at all frequencies tested and an average effect of –62.52 ± 13.13% (n = 5). Subsequent addition of atropine enhanced relaxation at 20 Hz (p < 0.05). Apamin fully blocked the l-NAME-resistant on relaxation in four of five experiments (p < 0.05). In the same experiment, l-NAME reduced the amplitude of basal EFS off contraction with significant effects at 0.5 to 20 Hz (p < 0.05). The subsequent addition of atropine further reduced off contraction at 10 and 20 Hz and apamin switched the off contraction to an on contraction during EFS without affecting the amplitude of the response. These results clearly show that EFS-induced LES relaxation is mainly mediated by both nitric oxide synthesis and other neurotransmitters acting through apamin-sensitive K+ channels. In the second experiment (Fig. 5, experiment E-2), we explored the possible effect of any inhibitory neurotransmitter released by nitric oxide. In this experiment, ODQ reduced EFS relaxation with significant effects at all frequencies tested (n = 5, p < 0.05). Subsequent addition of atropine increased the relaxation at 10 and 20 Hz, and apamin fully blocked this non-nitrergic EFS on relaxation in four of five experiments (p < 0.05). Also in experiment E-5, ODQ reduced the amplitude of the off contraction at frequencies above 1 Hz (p < 0.05); subsequent addition of atropine further reduced the amplitude of the contraction at 10 and 20 Hz (p < 0.05). Likewise, apamin switched the off response into an on contraction. In these experiments with stimulation of enteric motor neurons by EFS, the average of inhibitory effect of ODQ on relaxation (–67.67 ± 6.80%) was similar to that caused by l-NAME, arguing against the possibility of a relaxatory effect of any neurotransmitters released by nitric oxide. Apamin also caused similar and comparable effects in both experiments.
The nature of the non-nitrergic neurotransmitter released by EFS was explored in experiments E-3 to E.6. In experiment E-3 (Fig. 6), MRS 2179 strongly inhibited the non-nitrergic EFS relaxation with significant effects at 3 to 20 Hz (n = 5, p < 0.001). In experiment E-4, NF 279 slightly but consistently reduced non-nitrergic EFS relaxation, also with significant effects at 3 to 20 Hz (n = 5, p < 0.001). In experiment E-5 (Fig. 6), α-chymotrypsin only slightly inhibited the l-NAME-resistant on relaxation with significant effects at 3 and 20 Hz (n = 5, p < 0.05), and in experiment E-6, SnPP-IX did not affect EFS relaxation (n = 5, N.S.). Results from experiments E-1 suggest that EFS relaxation is mainly mediated by nitric oxide and an apamin-sensitive neurotransmitter, and results from experiments E-3 and E-4 show that the non-nitrergic and apamin-sensitive component is strongly inhibited by a purinergic P2Y1 receptor antagonist and only slightly reduced by a P2X1,2,3 purinergic receptor antagonist. This clearly indicates the effect of a purinergic neurotransmitter mainly acting at P2Y1 receptors through apamin-sensitive small conductance Ca2+-activated K+ channels. In addition, the non-nitrergic and apamin-sensitive component was only slightly reduced by α-chymotrypsin and unaffected by heme oxygenase inhibition, suggesting only a minor role for a peptidic neurotransmitter agreeing with the pharmacological profile of PACAP because the direct effect of PACAP on LES is inhibited by both apamin and α-chymotrypsin and no role for carbon monoxide during EFS relaxation.
Inhibitory Neurotransmitters in Nicotine-Induced Relaxation. Nicotine (1–300 μM) relaxed LES strips in a concentration-dependent manner. EC50 was 20.2 μM, and maximal relaxation (3.93 ± 1.03g or 100.36 ± 3.37% of resting tone) was reached at 100 μM (n = 5). The amplitude of the relaxation induced by stimulation of nicotinic AChRs was similar to that obtained by EFS (N.S.). Hexamethonium fully blocked nicotine-induced LES relaxation (–98.27 ± 3.51%, n = 3, p < 0.05), and tetrodotoxin reduced by 28.08 ± 12% but did not block maximal relaxation induced by nicotine (n = 5, p < 0.05).
Nicotine-induced relaxation (100 μM) was also significantly reduced by l-NAME by 60.37 ± 10.80% (n = 5, p < 0.05, Fig. 7, experiment N-1). Subsequent addition of apamin blocked non-nitrergic relaxation and transformed the nicotine response into a biphasic contraction with an amplitude of 0.65 ± 0.14 and 1.76 ± 0.19 g, respectively. This double-peaked contraction was fully blocked by atropine (p < 0.05). In parallel experiments (Fig. 7, experiment N-2), ODQ significantly reduced the nicotine-induced relaxation by 41.90 ± 7.98% (n = 5, p < 0.05), and apamin further blocked the nicotine relaxation, which also led to a cholinergic double-peaked contraction (0.77 ± 0.13 and 2.10 ± 0.41 g, respectively). In these experiments with stimulation of enteric motor neurons trough nicotinic AChRs, the effect of ODQ was similar to that induced by l-NAME (N.S.) and comparable with that caused by apamin in both parallel studies. Relative inhibition caused by l-NAME, ODQ, and apamin did not differ in both EFS and nicotine experiments (N.S.).
Experiments N-3 to N-6 (Fig. 8) were performed to characterize the nature of the non-nitrergic neurotransmitters released after stimulation of inhibitory LES motor neurons through nicotinic AChRs. Nicotine-induced relaxation resistant to l-NAME was strongly inhibited by MRS 2179 by 76.6 ± 9.27% (n = 5, p < 0.01) in experiment N-3 and also significantly inhibited by NF 279 by 48.04 ± 16.64% (n = 5, p < 0.05) in experiment N-4 and by 18.88 ± 1.00% in experiment N-5 by α-chymotrypsin (n = 5, p < 0.001) in experiment N-5. In experiment N-6, SnPP-IX did not significantly affect nicotine relaxation. Experiment N-1 shows that LES relaxation after stimulation of inhibitory motor neurons through nicotinic AChRs is mainly caused by nitric oxide and one or more apamin-sensitive neurotransmitters. The non-nitrergic and apamin-sensitive relaxation was strongly antagonized by the purinergic antagonist MRS 2179 in experiment N-3, further confirming the effect of a purinergic neurotransmitter mainly acting at P2Y1 receptors through apamin-sensitive small conductance Ca2+-activated K+ channels. In addition, non-nitrergic and apamin-sensitive nicotine-induced relaxation is also moderately antagonized through purinergic P2X1,2,3 receptor antagonists and by α-chymotrypsin, also suggesting a minor role for a peptidic neurotransmitter agreeing with the pharmacological profile of PACAP and no role for carbon monoxide in LES relaxation after stimulation of inhibitory motor neurons through nicotinic AChRs.
Discussion
Our study shows that porcine LES tone is mainly myogenic and is modulated by tonic input from excitatory and inhibitory enteric motor neurons. Relaxation of porcine LES after stimulation of inhibitory motor neurons is caused to a similar extent by two pathways: a nitrergic one mediated by nitric oxide through guanylate cyclase signaling and apamin-insensitive mechanisms and a non-nitrergic pathway coupled to apamin-sensitive small conductance Ca2+-activated K+ channels mainly mediated by ATP, ADP, or a related purine acting through P2Y1 receptors and a minor contribution of purinergic P2X1,2,3 receptors and PACAP. Our results suggest parallel release of nitrergic, purinergic, and peptidergic neurotransmitters, independent effects, and no major interplay on their release or postjunctional effects (Fig. 9). Our study did not find any role for CGRP in porcine LES, and although carbon monoxide and VIP caused a direct LES relaxation, we did not find any major role for them in neuroeffector LES relaxation.
Porcine, like human, LES is formed by the sling muscle on the angle of Hiss and the clasp component in lesser curvature (Preiksaitis and Diamant, 1997). We only included clasp strips that developed active tension and relaxed during stimulation of inhibitory motor neurons, ensuring that we had selected LES strips with intact innervation. We tested the effect of a nitric oxide donor, VIP, ATP, PACAP, a carbon monoxide donor, and CGRP because these substances colocalize in esophageal inhibitory motor neurons (Ny et al., 1995a; Uc et al., 1997). Colocalized substances can participate as neuromodulators, neurotransmitters, or with neurotrophic effects (Burnstock, 2004). We found that the nitric oxide donor, VIP, ATP, ADPβS, the carbon monoxide donor, and PACAP relaxed LES strips through a tetrodotoxin-insensitive mechanism, suggesting a direct effect on smooth muscle cells. Nitric oxide, ATP, and PACAP induced a complete LES relaxation, and CORM-1 and VIP were clearly less efficient, as observed in porcine gastric fundus (Colpaert et al., 2002). In contrast, the same dose of CGRP that relaxed LES strips in opossums failed to relax our preparation despite the similarity between both species in the EFS-induced responses (Uc et al., 1997). These discrepancies could be explained by species differences and/or CGRP acting through more complex neural circuits in “in vivo” studies (Rattan et al., 1988). Relaxation induced by sodium nitroprusside was specifically blocked by ODQ in our study, showing a direct nitric oxide effect through guanylate cyclase pathways (Shahin et al., 2000). VIP relaxation was unaffected by l-NAME or by ODQ arguing against the theory of a “serial cascade” involving nitric oxide production by VIP (Grider et al., 1992; Ergun and Ogulener, 2001). Apamin inhibited ATP and PACAP but not nitric oxide, CORM-1, or VIP relaxation; and α-chymotrypsin antagonized both VIP and fast-PACAP relaxation without affecting the nitric oxide or ATP response. ATP induced a complex LES response similar to that observed on vascular smooth muscle (Ralevic, 2002). Smooth muscle cells along the gastrointestinal tract can simultaneously express G-protein-coupled P2Y receptors mediating the relaxatory effects of purines and ion-gated P2X receptors mediating either contractions (Giaroni et al., 2002) or relaxations (Storr et al., 2000). Activation of P2X receptors could initially induce an increase in intracellular Ca2+ leading a contraction and a relaxation as a result of secondary activation of apamin-sensitive Ca2+-dependent K+-channels (Ishiguchi et al., 2000). MRS 2179 is a specific antagonist for P2Y1 purinoreceptor subtype (De Man et al., 2003) that in our study strongly antagonized the ATP relaxation. Neither l-NAME nor α-chymotrypsin affected ATP-induced relaxation, also arguing against an in-series relation between nitric oxide or VIP and ATP in our study (Xue et al., 2000).
We assessed the physiological relevance of nitric oxide, ATP, PACAP, carbon monoxide, and VIP by testing the effect of their antagonists on LES resting tone and during stimulation of inhibitory motor neurons by EFS or through nicotinic AChRs. Our results showed an intense reduction in LES resting tone by the Ca2+-free buffer, further confirming that porcine LES tone is mainly myogenic and depends on extracellular Ca2+ (Tottrup et al., 1990). Our study also showed that atropine decreased LES less intensively tone than the removal of extracellular Ca2+, showing a moderate tonic input from cholinergic motor neurons; and l-NAME, ODQ, and apamin enhanced LES tone, also showing a tonic influence of inhibitory motor neurons. Nitric oxide synthase inhibitors also increased LES resting tone in humans (Gonzalez et al., 2004). The higher effect of l-NAME compared with ODQ in our study could be explained by a finding in canine LES showing ODQ-dependent mechanisms mediating the action of nitric oxide from nerves, but not that from muscle (Daniel et al., 2002). Interestingly, we failed to show any effect of atropine on LES tone in an earlier study on human LES (Gonzalez et al., 2004), suggesting a stronger influence of cholinergic motor neurons in porcine LES that could affect treatments for achalasia based on the reduction of cholinergic inputs such as botulin toxin (Pasricha et al., 1993).
Porcine LES responses to EFS and nicotine in this study are similar to those we found in studies on human LES (Gonzalez et al., 2004). Our results on pigs show that maximal EFS relaxation was inhibited to the same degree by l-NAME and ODQ (by 60%) and that apamin fully blocked the EFS induced relaxation resistant to nitric oxide inhibition in both series of experiments. The apamin-sensitive component has been indirectly attributed to ATP in previous studies (Imaeda and Cunnane, 2003). We found that the apamin-sensitive component of EFS relaxation is strongly inhibited by MRS 2179, showing that a purine mediates this component by acting on P2Y1 receptors; it is slightly but consistently inhibited by NF 279, also suggesting the involvement of P2X1,2,3 receptors; and it is slightly inhibited by the peptidase chymotrypsin, suggesting that PACAP (an apamin-sensitive peptidergic neurotransmitters) also has a minor role in EFS-induced relaxation. Similar results on the involvement of P2Y1 and also P2X1,2,3 receptors on inhibitory neurotransmission were found in the mouse jejunum (De Man et al., 2003), and ATP-induced relaxation of rat pylorus has been attributed to muscular P2X purinoceptors (Ishiguchi et al., 2000). Heme oxygenase inhibitors did not modify the EFS-inhibitory responses in our preparation or in porcine stomach (Colpaert et al., 2002) or porcine ileum (Matsuda et al., 2004), precluding a major role for carbon monoxide. In addition, the similarity of the responses in experiments with ODQ (where nitric oxide is synthesized and released and could induce the release of any other neurotransmitter) and l-NAME (where nitric oxide is not synthesized) argues against release of VIP or any other neurotransmitter by nitric oxide (Grider et al., 1992) and also against nitric oxide inhibition of the release of ATP (Ishiguchi et al., 2000). Our results suggest mechanisms of simultaneous release and independent actions (cotransmission) for the nitrergic, purinergic, and peptidergic inhibitory neurotransmitters released during EFS.
We also explored the neurotransmitters released on stimulation of enteric motor neurons with nicotine. Nicotinic AChRs are located in somatodendritic regions of enteric motor neurons and participate in the transmission of vagal inputs (Galligan, 1999; Chang et al., 2003). We also found nicotinic AChRs at prejunctional sites of inhibitory motor neurons in human LES (Gonzalez et al., 2004). In the present study, nicotine-induced LES relaxations were partly antagonized by tetrodotoxin, further suggesting the presence of nicotinic AChRs in nerve terminals of the inhibitory neurons of porcine LES. Previous studies on cats (Kortezova et al., 1994) and our studies on human LES also found a residual relaxation induced by nicotine after nitric oxide blockade. Nitric oxide synthesis inhibitors reduced nicotine relaxation by more than 85% in humans (Gonzalez et al., 2004), 70 to 80% in cats (Kortezova et al., 1994), and by 40 to 60% in the present study on pigs. On the other hand, the nicotinic agonist 1,1-dimethyl-4-phenylpiperazinium induced both VIP and nitric oxide release, VIP release being further facilitated by nitric oxide production in the guinea pig ileum (Grider et al., 1992). In our present study, the relaxation resistant to nitric oxide blockade was fully antagonized by apamin and a weak contraction induced by stimulation of cholinergic neurons appeared. Similarly to the findings on EFS relaxation, we also found three components on apamin-sensitive nicotine LES relaxation, the main component being antagonized by MRS 2179 and mediated by a purine through P2Y1 receptors, a second minor component antagonized by NF 279 through P2X1,2,3 receptors, and a minor third component antagonized by α-chymotrypsin and also probably mediated by PACAP (Smid and Blackshaw, 2000) and unaffected by inhibitors of heme oxygenase. Our results also showed that LES relaxation induced by nicotine in basal conditions was similar to that induced by EFS, suggesting high efficiency of stimulation of inhibitory motor neurons by nicotinic AChRs. The off contraction induced by EFS in this study was partly antagonized by atropine, partly caused by a rebound depolarization because an important part of the off contraction is inhibited by nitric oxide blockers and partly caused by the probable participation of a secondary noncholinergic excitatory neurotransmitter such as tachykinins (Chang et al., 2003). In this experimental setting, stimulation of nicotinic AChRs during simultaneous blockade of nitrergic and apamin-sensitive pathways induced only a slight contraction in LES, suggesting that nicotinic AChRs stimulate excitatory motor neurons much less efficiently than inhibitory ones, as we found in human LES (Gonzalez et al., 2004). These results further suggest that although vagal fibers or interneurons could easily and efficiently stimulate inhibitory motor neurons, full stimulation of intrinsic excitatory enteric motor neurons requires non-nicotinic neurotransmitters (Galligan et al., 2000) or other circuits that need further investigation.
Acknowledgments
We thank Escorxador Frigorific Avinyo S.A. for the porcine tissue; Oscar Estrada, Anna Maria Alcántara, Antonio Acosta, and Susanna Comellas for technical support; and Marcel Jiménez and Jane Lewis for reviewing the manuscript.
Footnotes
- Received August 26, 2005.
- Accepted November 21, 2005.
-
This work was supported by the Fundació de Gastroenterologia Dr. Francisco Vilardell, by the Fundació Salut del Consorci Santari del Maresme, by the Ministerio de Sanidad y Consumo (FIS PI/020662), and by the Departament d'Universitats, Recerca i Societat de la Información (Grant SGR2001-0214).
-
This study was presented in part at the VIII Little-Brain Big-Brain Meeting in S'Agaró; 2003 October 1–5; Girona, Spain and in its final form at the 20th International Symposium on Neurogastroenterology and Motility; 2005 July 3–6; Toulouse, France.
-
doi:10.1124/jpet.105.094482.
-
ABBREVIATIONS: LES, lower esophageal sphincter; GEJ, gastroesophageal junction; VIP, vasoactive intestinal peptide; PACAP, pituitary adenylate cyclase-activating peptide; CGRP, calcitonin gene-related peptide; AChR, acetylcholine receptor; EFS, electrical field stimulation; l-NAME, l-NG-nitroarginine methyl ester; ODQ, 1H-[1,2,4]oxadiazole-[4,3-α]quinoxalin-1-one; CORM-1, tricarbonyl dichlororuthenum dimer; MRS 2179, N6-methyl 2′-deoxyadenosine 3′,5′-bisphosphate; NF 279, 8,8¢-[carbonylbis(imino-4,1-phenylenecarbonylimino-4,1-phenylenecarbonylimino)]bis-1,3,5-naphthalenetrisulfonic acid hexasodium salt; SnPP-IX, tin protoporphyrin IX; ADPβS, adenosine 5′-O-2-thiodiphosphate; DMSO, dimethyl sulfoxide.
- The American Society for Pharmacology and Experimental Therapeutics