In enteric synaptosomes of the rat, the role of voltage-dependent Ca2+channels in K+-induced VIP release and nitric oxide (NO) synthesis was investigated. Basal VIP release was 39 ± 4 pg/mg, and cofactor-substituted NO synthase activity was 7.0 ± 0.8 fmol · mg−1 · min−1. K+ depolarization (65 mM) stimulated VIP release Ca2+ dependently (basal, 100%; K+, 172.2 ± 16.2%; P < 0.05, n = 5). K+-stimulated VIP release was reduced by blockers of the P-type (ω-agatoxin-IVA, 3 × 10−8 M) and N-type (ω-conotoxin-GVIA, 10−6 M) Ca2+ channels by ∼50 and 25%, respectively, but not by blockers of the L-type (isradipine, 10−8 M), Q-type (ω-conotoxin-MVIIC, 10−6 M), or T-type (Ni2+, 10−6 M) Ca2+ channels. In contrast, NO synthesis was suppressed by ω-agatoxin-IVA, ω-conotoxin-GVIA, and isradipine by ∼79, 70, and 70%, respectively, whereas Ni2+ and ω-conotoxin-MVIIC had no effect. These findings are suggestive of a coupling of depolarization-induced VIP release primarily to the P- and N-type Ca2+ channels, whereas NO synthesis is presumably dependent on Ca2+ influx not only via the P- and N- but also via the L-type Ca2+ channel. In contrast, none of the Ca2+ channel blockers affected VIP release evoked by exogenous NO, suggesting that NO induces VIP secretion by a different mechanism, presumably involving intracellular Ca2+ stores.
- enteric nervous system
- vasoactive intestinal polypeptide
- voltage-dependent Ca2+ channels
- nitric oxide synthase
the activation ofvoltage-dependent Ca2+ channels (VDCCs) is a key regulatory step in the process of excitation-secretion coupling (4, 25). Molecular neurobiology has revealed a diversity of Ca2+ channel subtypes (for review, see Refs.15 and 22). They are classified according to their α1-subunit, which is encoded by seven different genes, and subdivided into classes A (P/Q-type), B (N-type), C and D (L-type), S (L-type), G (T-type) and E (R-type) Ca2+ channels (15). Cumulative evidence has been presented that multiple VDCCs may coexist in a single synapse to regulate neurosecretion from peripheral autonomic (31) and enteric nerves (37,41, 43). Immunohistochemistry revealed α1B and α1A channel-like immunoreactivity in nerve processes within the myenteric plexus of the rat small intestine, suggesting the presence of N- and P-type Ca2+ channels, respectively (26). This pattern of distribution closely resembles the localization of both channels in the central nervous system (34,38). An α1C immunoreactivity, implying the presence of L-type Ca2+ channels, appeared to be more confined to ganglionic structures than to nerve processes (26). Furthermore, binding sites for both ω-conotoxin-GVIA (ω-Ctx GVIA), a blocker of the N-type Ca2+ channel, and nitrendipine, a blocker of the L-type channel, had been demonstrated on isolated synaptosomes of the canine small intestine (1), suggesting a role of both channel types in neurosecretion. However, the functional relevance of these findings remains to be elucidated.
VIP represents an inhibitory nonadrenergic, noncholinergic (NANC) neurotransmitter in the gastrointestinal tract that acts in concert with nitric oxide (NO) to relax smooth muscle. Both VIP and the NO-synthesizing enzyme neuronal NO synthase (nNOS) are colocalized in a subset of descending inhibitory neurons in the rat (2, 17). Previously, it has been shown that enteric synaptosomes release VIP in response to exogenous NO and in a Ca2+-dependent manner following stimulation with KCl (3) and that a constitutive Ca2+-dependent NOS activity is retained within isolated nerve terminals from rat small intestine (23), which is functionally coupled to VIP release (29). VIP output from longitudinal muscle preparations with adherent myenteric plexus (LM/MP) has been shown to be reduced by ω-Ctx GVIA, a blocker of the N-type Ca2+channel (11, 13). Since NO is known to relax smooth muscle directly or indirectly by releasing VIP (3, 21), and since the effect of NO on smooth muscle is antagonized in the presence of ω-Ctx GVIA (6, 11, 13), it was of interest to investigate the effect VDCC blockers on NO-induced VIP release as well. Furthermore, it must be clarified whether selective blockade of VDCCs other than the N-type channel elicits an effect on VIP release or NO synthesis.
On the ultrastructural level, NOS appears not to be located in the same domain of Ca2+ entry as the VIP-containing vesicles (5), leaving the possibility that both transmitters are controlled independently. So far, no data have been presented regarding the involvement of distinct Ca2+ channels in both the release of VIP from and the control of NO synthesis within nerve terminals.
The aim of the present study was 1) to identify the types of Ca2+ channels regulating the release of VIP from enteric nerve terminals by studying Ca2+ channel blockers targeting subsets of N-type (ω-Ctx GVIA), L-type (isradipine), T-type (Ni2+), and P/Q-type [ω-agatoxin-IVA (ω-Agatx IVA) and ω-conotoxin-MVIIC (ω-Ctx MVIIC)] Ca2+ channels and2) to study the effect of the respective Ca2+channel blockers on NO production by nNOS. These questions have been addressed (28) by studying isolated nerve terminals that offer the unique opportunity to examine intracellular and subcellular mechanisms of neurotransmitter release without the interference of other local or systemic factors present in vivo or in the intact organ in vitro.
MATERIALS AND METHODS
Tissue handling and membrane preparation.
Synaptosomes were prepared as described previously (28). Briefly, five male Wistar rats were killed by cervical dislocation, and the small intestine was quickly removed and suspended in ice-cold buffer (20 mM MOPS, 10 mM MgCl2, 8% wt/vol sucrose, pH 7.4). All further preparative steps were carried out at 0–4°C. Approximately 6- to 8-cm pieces of small intestine were dissected, cleaned of mesenteric arcade and fat, and opened along the mesenteric attachment line. The mucosal layer was scraped off with a sharp razor blade, and the remaining muscle layers were put into cold buffer. The muscle tissue was blotted dry on filter paper and weighed. For membrane preparation, the tissue was resuspended in isolation buffer (8% wt/vol sucrose, 20 mM MOPS, pH 7.4), minced with scissors, and homogenized with a Polytron PT20 homogenizer at ∼1,500 rpm setting for 15 s (3 × 5 s).
Fractionation of tissue homogenate by differential centrifugation.
The tissue homogenate was centrifuged in two steps of 800 gfor 10 min to remove myofibrils and remaining nuclei. The supernatant was collected [postnuclear supernatant (PNS)] and recentrifuged at 3,500 g for 10 min to obtain the P1 fraction. The supernatant was centrifuged again at 100,000 g for 90 min. The pellet from this centrifugation [microsomal 1 (MIC1)] was resuspended and centrifuged again at 10,000 g for 10 min. The resulting pellet and the supernatant are referred to as mitochondrial 2 (P2) and microsomal 2 (MIC2) fractions, respectively.
Differential centrifugation led to a substantial enrichment of [3H]saxitoxin in the fraction P2 [8-fold (44.9 ± 8 fmol/mg protein) vs. PNS (5.5 ± 1.7 fmol/mg protein)] and was paralleled by a 7.2-fold increase in the content of VIP in the P2 fraction (6.4 ± 1.9 vs. 0.9 ± 0.3 pmol/mg protein in PNS), as reported previously (3). With respect to the methodical approach employed, the synaptosomal fraction consists of a mixed population of nerve terminals, including nerve endings of motoneurons, from different animals.
Protein was measured spectrophotometrically according to the method of Bradford (8). Bovine serum γ-globulin was used as standard.
VIP immunoreactivity was determined as described elsewhere (35). The porcine VIP antibody (provided by S. R. Bloom, Royal Postgraduate Medical School, London, UK) showed no interaction with NH2-terminal fragments of VIP, secretin, peptide histidine isoleucine, growth hormone-releasing factor, gastric inhibitory peptide, or pituitary adenylate cyclase-activating peptide (personal communication, S. R. Bloom). 125I-VIP for the preparation of the labeled VIP and synthetic VIP as standard were purchased from Amersham and Sigma (Munich, Germany), respectively.
Peptide release studies were carried out in Krebs-Ringer-bicarbonate solution (KRS; in mM: 115.5 NaCl, 1.16 MgSO4, 1.16 NaH2PO4, 11.1 glucose, 21.9 NaHCO3, 2.5 CaCl2, 4.16 KCl) gassed with 95% O2-5% CO2. In experiments designed to study the role of Ca2+-free medium, CaCl2 was omitted and 0.5 mM EGTA was added. KRS (1,050 μl) and 150 μl of drugs or KRS alone serving as a control (basal level) were incubated in separate test tubes at 37°C in a gently shaking water bath. The reaction was started by adding 300 μl of synaptosomal membranes (300 μg protein) to each tube at timed intervals. To study the time course of Ca2+ channel blockade, the blockers were incubated with the synaptosomes for 0, 5, 15, and 20 min. According to the data obtained from the time-course experiments, the incubation for all subsequent experiments lasted 15 min. To stop the reaction, the synaptosomal membranes were put on ice and immediately sedimented by high-speed centrifugation in a refrigerated centrifuge. The supernatant was withdrawn and immediately frozen at −20°C until peptide determination by radioimmunoassay.
Assay of NOS activity.
NOS activity was determined by monitoring the formation ofl-citrulline from l-arginine by a modification of methods described previously (9). Enzymatic reactions were conducted at 37°C in 25 mM MOPS-8% sucrose buffer containing 1 mM NADPH, 0.1 μM tetrahydrobiopterin, 1 μM calmodulin, 1 mM CaCl2, 0.1 μM FAD, 0.1 μM flavin mononucleotide (FMN), and other test agents as specified later, in a final incubation volume of 750 μl. Thel-[3H]arginine was purified by anionic exchange chromatography on columns of Dowex 1×8, OH-form to remove traces of l-[3H]citrulline. After preincubation for 30 min at 4°C with synaptosomes (500 μl), the enzymatic reactions were started by adding ∼500,000 dpm ofl-[3H]arginine (63 Ci/mmol) and terminated after 15 min by immediate heating to 90°C for 6 min and addition of 1 ml distilled water containing 1 mM l-arginine and 1 mMl-citrulline. Samples were applied to columns containing 1 ml of Dowex AG 50 W × 8 resin, Na+ form, preequilibrated with sucrose MOPS buffer. The eluate (2 ml) was collected in a liquid scintillation vial. After the addition of 1.5 ml scintillation fluid, samples were counted in a Beckman LS 3801 spectrometer. The recovery ofl-[3H]citrulline in the first 4 ml of the eluate was ∼92%; contamination withl-[3H]arginine did not exceed 2%. Basal values were obtained by heating samples to 100°C for 5 min before incubation. Cofactor-stimulated NOS activity over basal was considered as control. To study the effect of Ca2+-free medium, CaCl2 was omitted and 0.5 mM EGTA was added.
The labeled l-[3H]arginine (63 Ci/mmol) was purchased from Amersham. All other reagents were purchased from the indicated sources as follows: NADPH, FAD, FMN, citrulline, calmodulin, pepstatin A, dithiothreitol, trypsin inhibitor, Ni-sulfate, glutamate, carbachol, atropine, [d-penicillamine(2,5)]enkephalin (DPDPE), substance P, and naloxone hydrochloride dihydrate were from Sigma; tetrahydrobiopterin was from ICN Biomedicals (Eschwege, Germany);S-nitroso-N-acetylpenicillamine (SNAP) andN-methyl-d-aspartate (NMDA) were from Calbiochem (Bad Soden, Germany); thiorphan was from Fluka (Munich, Germany); R-4-carboxy-3-hydroxyphenylglycine, 1-aminocyclobutanecarboxylic acid (ACBC), and endomorphin-1 were from Tocris-Cookson Biotrend Chemikalien (Cologne, Germany); [d-Lys(nicotinoyl)1,β-(3-pyridyl)-Ala3,3,4,dichloro-d-Phe5-Asn6-d-Trp7,9,Nle11]-substance P (Spantide II) was from Bachem (Heidelberg, Germany); U-50488 was from Upjohn (Kalamazoo, MI); BAY K 8644 and diethylamine/NO complex sodium (DEA-NO) was from Research Biochemicals International (Natick, MA); PMSF was from Serva (Heidelberg, Germany); ω-Agatx IVA was from Alexis (Grünberg, Germany); ω-Ctx GVIA and MVIIC were from Alomone Labs (Jerusalem, Israel); and isradipine was generously provided by Prof. F. Hofmann, (Dept. of Pharmacology, Technical University of Munich). All experiments with isradipine were carried out under light protection. Adequate controls were performed with the vehicles used for solubilizing each reagent.
Data are given as means ± SE; n indicates the number of independent observations in separate experiments from separate preparations. For each value of a given drug of a single preparation, the release study was carried out in duplicate. The values of peptide release experiments showed some variation in separate experiments and were therefore expressed as the relative increase over basal levels (=100%). Analysis of variance, followed by Dunnett's post hoc test for multiple testings, was used to determine statistical significance. For comparisons of two means, paired or unpairedt-test was performed. Values of P ≤ 0.05 were considered significant.
Assessment of VIP Release From Isolated Synaptosomes
Synaptosomes of rat small intestine are capable of releasing VIP in response to K+ depolarization. Depolarization-induced VIP release occurred in a Ca2+-dependent manner (Fig.1). Basal release was also significantly reduced (−40%) when incubation was carried out in a Ca2+-free medium containing 0.5 mM EGTA. No significant difference was noted between basal and K+-induced VIP release in the presence of Ca2+-free medium, suggesting that Ca2+-independent release is not likely to occur. A first set of experiments was conducted to investigate the effect of the respective selective Ca2+ channel blockers on basal VIP release. The inhibitors were used at concentrations selective for N-type (ω-Ctx GVIA, 10−6 M), N/Q-type (ω-Ctx MVIIC, 10−6 M), P-type (ω-Agatx IVA, 3 × 10−8 M), L-type (isradipine, 10−8 M), and R/T-type (Ni2+, 10−6 M) Ca2+channels. Of the blockers tested, only ω-Agatx IVA significantly inhibited basal VIP release to an extent not significantly different from Ca2+-free medium [basal, 36.8 ± 3.9 pg/mg (100%); ω-Agatx IVA, 86.5 ± 8.8%; P < 0.05;n = 6; Fig. 2]. None of the other blockers affected basal VIP release (basal, 100%; ω-Ctx GVIA, 122.1 ± 12.1%; ω-Ctx MVIIC, 140.6 ± 23.6%; isradipine, 113.8 ± 10.9%; Ni2+, 92.1 ± 16.4%).
Effect of Subtype-Specific Ca2+Channel Blockers on K+-Induced VIP Release
The sensitivity of K+-evoked release to Ca2+ channel blockade was studied. A first set of experiments was conducted to inquire about the time course of Ca2+ channel blockade by the blockers of the N-, P-, and L-type channels, ω-Ctx GVIA, ω-Agatx IVA, and isradipine, respectively. The substances ω-Ctx GVIA and ω-Agatx IVA time-dependently blocked K+-induced VIP release with a maximum effect at an incubation time of 15 min (Fig.3), whereas isradipine turned out to be ineffective.
In a second series of experiments using the 15-min incubation time, K+ depolarization-induced VIP release was significantly reduced by ∼25 and 50% in the presence of the N- and P-type channel blockers, ω-Ctx GVIA and ω-Agatx IVA, respectively [basal, 183.1 ± 3.3 pg/mg (100%); K+, 168.6 ± 7.2%; K+ + 10−6 M ω-Ctx GVIA, 152.0 ± 5.4% (P < 0.05; n = 5); K+ + 3 × 10−8 M ω-Agatx IVA, 135.2 ± 13.1% (P < 0.05, n = 5); Fig. 4], suggesting that both channels are coupled to VIP release evoked by K+. The combination of both blockers resulted in a further inhibition (10−6 M ω-Ctx GVIA + 3 × 10−8 M ω-Agatx IVA, 119.6 ± 15.7%; n = 5), which was not significantly different from that of ω-Agatx IVA alone.
The Q-type channel blocker ω-Ctx MVIIC had no significant effect [basal, 183.1 ± 3.3 pg/mg (100%); K+, 168.6 ± 12.2%; K+ + 10−6 M ω-Ctx MVIIC, 163.8 ± 17.5%; n = 5]. The L-type Ca2+ channel blocker isradipine (10−8M) did not suppress K+-evoked VIP release, nor did it augment the inhibition in the presence of ω-Ctx GVIA and ω-Agatx IVA [basal, 183.1 ± 3.3 pg/mg (100%); K+, 168.6 ± 7.2%; K+ + isradipine, 141.3 ± 9.7%; K+ + ω-Ctx GVIA + ω-Agatx IVA, 119.6 ± 15.7%; K++ ω-Ctx GVIA + ω-Agatx IVA + isradipine, 136.4 ± 21.5%; n = 5], suggesting that blockade of the L-type Ca2+ channel had no significant effect.
BAY K 8644 (10−6 M), which selectively activates L-type channels, failed to significantly stimulate either basal or K+-induced VIP release, respectively [basal, 149.8 ± 32.9 pg/mg (100%); BAY K 8644, 98.6 ± 20.8%; K+, 166.7 ± 18.4%; K+ + BAY K 8644, 134.8 ± 24.9%; n = 4].
Low-threshold T-type Ca2+ channels and high-threshold R-type Ca2+ channels are sensitive to blockade by Ni2+. Addition of Ni2+(10−6–10−4 M) had no significant effect on synaptosomal VIP release evoked by K+ [basal, 217.1 ± 3.3 pg/mg (100%); K+, 134.0 ± 4.3%; 10−6 M Ni2+, 125.3 ± 26.6%; 10−5 M Ni2+, 159.1 ± 34.9%; 10−4 M Ni2+, 141.9 ± 30.9%;n = 3].
Since the synaptosomal fraction consists of a mixed population of enteric nerves, which could release their neurotransmitter in response to K+ depolarization, the observed effect of the VDCCs could be due to an indirect action on other nerve terminals than the VIPergic one. To clarify this, we investigated whether carbachol, substance P, or the opioid peptides had a stimulatory effect on VIP release from nerve terminals. The respective data are shown in Table1.
Effect of Subtype-Specific Ca2+Channel Blockers on NOS Activity in Enteric Synaptosomes
NOS activity (in fmol · mg−1 · min−1) was decreased by 61% in the absence of Ca2+ [control (all cofactors included over basal), 7.0 ± 0.8; Ca2+-free, 2.8 ± 1.4; P < 0.05; n = 6]. Either treatment with ω-Agatx IVA, ω-Ctx GVIA, or isradipine reduced NOS activity compared with control (control, 7.0 ± 0.8; ω-Agatx IVA, 1.5 ± 1.2; ω-Ctx GVIA, 2.1 ± 1.3; isradipine, 2.1 ± 1.2; either treatment P < 0.05; n = 5; Fig. 5). Each of the three substances significantly reduced prestimulated NOS activity, corresponding to an inhibition by 79, 70, and 70%, respectively. No significant difference was obtained when these treatments were compared with Ca2+-free medium. In contrast, ω-Ctx MVIIC or Ni2+ (control, 7.0 ± 0.8; 10−6 M ω-Ctx MVIIC, 3.4 ± 1.8; 10−6 M Ni2+, 4.4 ± 1.9; n = 5) failed to significantly reduce cofactor-stimulated NOS activity (Fig. 5). Higher concentrations of Ni2+ were also not effective (control, 3.1 ± 1.3; 10−6 M Ni2+, 3.7 ± 1.6; 10−5 M Ni2+, 3.6 ± 1.7; 10−4 M Ni2+, 3.1 ± 1.2;n = 4).
These data suggest that, with respect to NO synthesis, not only P- and N- type channels but also L-type channels are involved, whereas Q- and T-type channels are presumably of minor relevance. Since in the central nervous system NMDA receptor activation induces Ca2+influx, thereby stimulating NOS, the effect of both NMDA receptor stimulation and blockade was studied. Addition of NMDA or glutamate does not alter cofactor-stimulated NOS activity (control, 2.2 ± 0.1; 10−7 M NMDA, 2.0 ± 0.3; 10−6 M NMDA, 1.9 ± 0.3; 10−5 M NMDA, 1.8 ± 0.4; 10−7 M glutamate, 2.1 ± 0.3; 10−6 M glutamate, 2.0 ± 0.3; 10−5 M glutamate, 1.9 ± 0.2; no treatment was significant vs. control; n = 5). The NMDA receptor antagonist R-4-carboxy-3-hydroxyphenylglycine also did not modify NOS activity (control, 2.2 ± 0.1; 10−6 M R-4-carboxy-3-hydroxyphenyl-glycine, 2.0 ± 0.3; not significant vs. control; n = 4). Similar results were obtained with ACBC, an NMDA receptor antagonist at the glycine site (control, 2.2 ± 0.1; 10−6 M ACBC, 2.9 ± 1.1; not significant vs. control; n = 4).
Since addition of Ca2+ evokes the release of other neurotransmitters, which in turn might stimulate NO production in synaptosomes, a new series of experiments was conducted. The data obtained with carbachol, substance P, and the respective agonists at the opioid μ-, δ-, and κ-receptors, endomorphin-1, DPDPE, and U-50488 are shown in Table 2.
Effect of Subtype-Specific Ca2+-Channel Blockers on VIP Release by Exogenous NO
SNAP induced a significant release of VIP, which was blocked in the presence of the NO scavenger oxyhemoglobin (basal, 100%; 10−4 M SNAP, 200.4 ± 22.5%; SNAP + 10−3 M oxyhemoglobin, 112.4 ± 21.2%;P < 0.01; n = 7), indicating that free NO is involved. Oxyhemoglobin (10−3 M) itself did not alter basal VIP release.
Since the foregoing data suggest the presence of N- and P-type Ca2+ channels on enteric synaptosomes and their functional relation to VIP secretion, another series of experiments was conducted to inquire about the role of VDCCs in VIP release under the conditions employed with exogenous NO.
In Ca2+-free medium containing 0.5 M EGTA, NO-induced VIP release was not significantly reduced [basal, 135.2 ± 26.2 pg/mg (100%); 10−4 M SNAP, 160.4 ± 15.1%; SNAP/Ca2+-free medium, 131.2 ± 27.9%;n = 6]. Accordingly, neither treatment with any subtype-selective Ca2+ channel blocker was able to significantly antagonize NO-induced VIP release [basal, 168.0 ± 24.7 pg/mg (100%); 10−4 M SNAP, 149.4 ± 11.4%; SNAP + 10−6 M ω-Ctx GVIA, 153.1 ± 12.4%; SNAP + 3 × 10−8 M ω-Agatx IVA, 150.6 ± 12.5%; SNAP + ω-Agatx IVA + ω-Ctx GVIA, 162.2 ± 25.6%; SNAP + isradipine, 142.4 ± 10.9%; SNAP + ω-Ctx MVIIC, 163.6 ± 27.7%; SNAP + Ni2+, 180.4 ± 42.2%; n = 7]. An additional series of experiments with the NO donor DEA-NO (10−6 M), a NONOate, confirmed the previous data obtained with SNAP (data not shown).
NANC relaxation of gastrointestinal muscle is mediated by NO and VIP, both of which are localized in descending inhibitory interneurons of the enteric nervous system. With respect to smooth muscle relaxation, interaction between these two transmitters appears to be likely; however, it is hard to assess since the enteric nervous system is a complex system and interference of local or systemic factors cannot be ruled out in vivo or in the intact organ in vitro. As an alternative approach, we used a synaptosomal preparation to inquire about the subcellular mechanisms involved in the release of VIP and NO synthesis. With respect to VIP release, previous data from enteric nerve terminals (3) and from isolated ganglia (21) could be confirmed showing that K+ is capable stimulating VIP release Ca2+ dependently. NO synthesis also strictly requires Ca2+ (23). Additionally, it could be shown that within isolated enteric nerve terminals NO synthesis is both coupled to VIP release (29) and under feedback control of endogenous or exogenous NO (30). There is evidence that NO-mediated relaxation of intestinal smooth muscle is strongly dependent on Ca2+entry through N-type Ca2+ channels, which are blocked specifically by ω-Ctx GVIA (7, 11, 13). Accordingly, experiments were performed to investigate the role of different VDCCs in the release of VIP and NO synthesis in enteric synaptosomes. Immunohistochemistry revealed the presence of the α1A- and α1B-subunits corresponding to the P- and N-type Ca2+ channels as well as the α1C- and α1D-subunits corresponding to the L-type Ca2+channel in the enteric nervous system, respectively (26). The data presented here clearly indicate that, following K+depolarization, N- and P-type Ca2+ channels are apparently coupled to VIP release, which extends previous findings in LM/MP preparations (13). Basal VIP output presumably is under control of the P-type Ca2+ channel. However, the synaptosomal fraction contains a mixed population of nerve endings, releasing their respective neurotransmitters in the presence of depolarizing stimuli. Assuming that some of these nerve endings synthesize a transmitter that stimulates VIP release, it could be speculated that the VDCCs might act indirectly and not at VIPergic nerve terminals. In another series of experiments, we inquired about a possible effect of other transmitters on VIP release. The “chemical coding” of the rat gut is quite different from other species (18). Inhibitory motor neurons contain almost exclusively NOS and/or VIP. The transmitters of excitatory motor neurons are acetylcholine, substance P, and/or opioid peptides, whereas the secretomotor neurons contain either VIP or substance P (18). Other peptidergic transmitters, detected by immunohistochemistry in the rat enteric nervous system, like somatostatin, neuromedin U, neurotensin, and CGRP, are almost exclusively confined to enteric ganglia. Within nerve fibers supplying the smooth muscle, they are observed in small numbers only (16). Enteric ganglia, however, are separated by the means of differential centrifugation during the preparation of synaptosomes (28). With respect to recently published functional data showing that both galanin and neuropeptide Y, which are present in moderate numbers within nerve fibers to smooth muscle (16), inhibit VIP output from the LM/MP preparation (12), the transmitters substance P, acetylcholine, and opioid peptides appeared to be the most likely candidates to stimulate VIP release. Our data show that none of the latter transmitters has a stimulatory effect on VIP release, suggesting that the VDCCs are acting directly on VIPergic nerves.
A second series of experiments was conducted to investigate the role of Ca2+ channels in NO production. Our data confirmed and extended previous findings (13) showing that not only N-type but also L-type and P-type Ca2+ channels provide Ca2+ influx into the nerve terminal to enable NO production. In analogy to the VIP assay, we investigated whether the effect is due to an indirect modulatory effect of VDCCs on nerve terminals containing either substance P, acetylcholine, or opioid peptides, which in turn could stimulate NO production. The latter substances did not modify NOS activity, suggesting a direct modulatory effect of the VDCCs on NOS-containing neurons.
Interestingly, NO biosynthesis in the brain is regulated by NMDA receptor-activated Ca2+ influx (19, 20). In enteric synaptosomes, both NMDA/glutamate and the NMDA receptor antagonists failed to influence NO synthesis significantly. These results indicate that distinct Ca2+ influx pathways specifically regulate nNOS activity in various tissues.
Further experiments were conducted to clarify the role of VDCCs in NO-mediated VIP release. In contrast to depolarization-induced VIP release, the NO-induced VIP secretion was not blocked in the presence of Ca2+ channel blockers. There is cumulative evidence that NO triggers exocytosis by bypassing requirements of Ca2+influx (33, 40). NO could accordingly release VIP in different ways, interacting with the release machinery (32), modulating ion channels on the membrane (27), or activating intracellular signal transduction systems (42). The last of these has been demonstrated for NO-induced VIP release from enteric synaptosomes, since it could be antagonized in the presence of blockers of the cGMP/protein kinase G pathway (29). Previous studies in the central nervous system have shown that cGMP analogs and NO donors reduce synaptosomal [Ca2+]i (33), which appears to be incompatible with the generally accepted Ca2+requirement for exocytosis. Our data support the notion that VIP release, although normally highly Ca2+-dependent, might occur in a Ca2+-independent manner. Further experiments are needed to clarify the exact mechanism involved.
From the data presented, clear evidence could be obtained that the suppression of NO-mediated intestinal relaxation by the N-type Ca2+ channel antagonist ω-Ctx GVIA (6, 13) is presumably not due to inhibition of NO-induced VIP release. The data also show that NO production and VIP release by depolarizing stimuli are differentially regulated with respect to the involvement of L-type channels. The functional role of these L-type channels in NO synthesis needs to be clarified. The L-type channel is encoded by the α1C- and α1D-subunits. The presence of both α1C- and α1D-subunits in enteric nerve terminals had been demonstrated by immunochemistry (26). L-type channels apparently are involved in neurotransmitter release from cerebellar synapses (36), NANC nerves supplying the iris muscle (24), and dendrites of hippocampal granule cells (39), but its involvement in the release process of neuropeptides in the enteric nervous system has not yet been established. Another concept for the functional role of L-type channels in neuronal tissue is provided by studies in hippocampal neurons. It could be shown that mobilization of calmodulin from cytosolic sources can be induced by certain Ca2+ entry systems, like L-type channels or NMDA receptors, but not by N- or P-type channels (14). Since binding of calmodulin to NOS is prerequisite for NO production and nNOS itself is targeted to specific intraterminal regions (10), the mobilization of calmodulin could be functionally related to the L-type channel. However, this is speculative and deserves further elucidation.
In conclusion, we demonstrated that VIP release following K+ depolarization is coupled to N- and P-type Ca2+ channels, whereas NO synthesis is presumably dependent on Ca2+ influx via the N-, P-, and L-type Ca2+channels. However, VDCCs appear not to be functionally coupled to the VIP release in response to extracellular NO, implying the possibility that intracellular Ca2+ stores, for example of mitochondrial origin, are of central importance for an interplay between the two messenger substances NO and Ca2+.
We acknowledge the cooperation of Prof. Dr. F. Hofmann, Department of Pharmacology, Prof. Dr. B. Gänsbacher, Prof. Dr. W. Erhardt, and their collaborators, Department of Experimental Surgery, Technical University of Munich. We also thank Heidi Paeghe and Sabine Herda for their expert technical assistance.
This study was supported by Deutsche Forschungsgemeinschaft (SFB 391/C5).
Parts of this work have been presented in abstract form at the annual meeting of the American Gastroenterological Association (New Orleans, 1998).
Address for reprint requests and other correspondence: M. Kurjak, Dept. of Internal Medicine II, Technical Univ. of Munich, Ismaninger Str. 22, 81675 Munich, Germany (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
July 31, 2002;10.1152/ajpgi.00400.2001
- Copyright © 2002 the American Physiological Society