Roles of Guanylate Cyclase in Responses to Myogenic and Neural Nitric Oxide in Canine Lower Esophageal Sphincter

Transcription

1 /02/ $7.00 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 301, No. 3 Copyright 2002 by The American Society for Pharmacology and Experimental Therapeutics 4599/ JPET 301: , 2002 Printed in U.S.A. Roles of Guanylate Cyclase in Responses to Myogenic and Neural Nitric Oxide in Canine Lower Esophageal Sphincter E. E. DANIEL, 1 T. J. BOWES, and J. JURY Department of Medicine, McMaster University, Hamilton, Ontario, Canada Received October 2, 2001; accepted February 19, 2002 This article is available online at ABSTRACT Whether cgmp and cytosolic guanylate cyclase (cgc) mediate responses of canine lower esophageal sphincter (LES) to nitric oxide (NO) released from nerves, produced in muscle, or added exogenously was evaluated in vitro. 1-H-(1,2,4)oxadiazole(4,3- )quinoxalin-1-1 (ODQ), inhibitor of cgc, reduced relaxations to nerve stimulation and sodium nitroprusside but not to nitricoxide synthase activity-dependent outward K -currents in isolated muscle cells. ODQ also failed to increase tone after nerve blockade. Nonspecific K channel blocker, TEA ion at 20 mm was previously shown to increase tone, occlude NO-mediated Previously we showed that canine LES contains a membrane-bound, constitutive nitric-oxide synthase (Salapatek et al., 1998b,c) that acts to limit development of tone in canine LES. This nitric-oxide synthase, which appears to be neural (Daniel et al., 2001a), uses Ca 2 entering through nearby L-type calcium channels for activation and tone regulation by NO formation that activates K channels, including BK Ca channels (Daniel et al., 2000). However, the source is not neural as tetrodotoxin alone or with -conotoxin (GVIA) had no effect on tone (Salapatek et al., 1998a,c). Moreover, tetrodotoxin alone but not -conotoxin (GVIA) alone abolished relaxations to electrical field stimulation (Daniel et al., 2000). Another study showed that contraction in LES was supported by two extracellular sources of Ca 2, one of which supported spontaneous tone whereas the other, which supported contraction to carbamyl choline (carbachol), was resistant to removal by extracellular fluid with 0 Ca 2 and 100 M EGTA and appeared to recycle between extracellular binding sites and nearby Ca 2 stores (Salapatek et al., 1998a). Similar results have been obtained in airway smooth muscle (Montano et al., 1993, 1996; Bazan-Perking et al., 1998). In these studies, Ca 2 from both of these sources entered through L-type calcium channels. We suggested that Supported by the Medical Research Council, Canada. 1 Current address: Department of Pharmacology, University of Alberta, 9-70 Medical Sciences Building, Edmonton, AB T6G 2H7, Canada. modulation of tone, and inhibit NO-dependent outward currents but not neural relaxation in LES cells. In this study, TEA abolished neural relaxation and nearly abolished relaxation to sodium nitroprusside when present with ODQ. We conclude that mechanisms coupling NO in canine LES to responses vary with the source of NO. ODQ-dependent mechanisms, presumably involving cgc, mediate actions of NO from nerves, but NO from muscle utilizes TEA-sensitive but not ODQ-dependent mechanisms to modulate tone and outward currents. Exogenous NO utilizes both TEA- and ODQ-dependent mechanisms. the Ca 2 bound extracellularly was located in membrane caveolae, which also contained neuronal nitric-oxide synthase and L-type calcium channels based on studies in another canine smooth muscle and recently in LES (Darby and Daniel, 2000; Daniel et al., 2001a). We showed (Daniel et al., 2000) that relaxation by NO released from LES enteric nerves was unaffected by the same potassium channel blocking agents that inhibited NO-mediated muscle relaxation and outward currents. These potassium channel blocking agents, however, partially inhibited relaxations to exogenous NO whether delivered from sodium nitroprusside or 3-morpholinosydnonimine. NO-mediated LES relaxation was also unaffected by chlorine channel blockade. However, a combination of TEA and 4,4 -diisothiocyanostilbene-2,2 -disulfonic acid abolished nerve-mediated relaxations. These findings raised the question: are the same second messengers involved in signaling the action of NO when it is derived from nerves, from muscle, or added exogenously? NO signaling has been studied in many tissues and often involves activation of cytosolic guanylate cyclase with elevation of cgmp, activation of protein kinase G, and subsequent phosphorylation of various membrane proteins, as exemplified in recent references (Ignarro et al., 1999; Chang et al., 2000; Gorodeski, 2000a,b; Janssen et al., 2000; Kwan et al., 2000; Tseng et al., 2000; Yao et al., 2000). However, an increasing number of NO-mediated biological ABBREVIATIONS: LES, lower esophageal sphincter; NO, nitric oxide; cgc, cytosolic guanylate cyclase; EFS, electrical field stimulation; ODQ, 1-H-(1,2,4)oxadiazole(4,3- )quinoxalin-1-1; L-NAME, N -nitro-l-arginine methyl ester; ICC, interstitial cells of Cajal; pps, pulses per second; BK Ca channels, large conductance Ca 2 -dependent K channels; L-NOARG, N-nitro-L-arginine; ANOVA, analysis of variance. 1111

2 1112 Daniel et al. Fig. 1. Effects of 10 5 M ODQ on tone (compared with tone prior to adding ODQ, both expressed relative to original tone of 100%). Note that ODQ increased tone significantly above 100% when nerves were functional (top panel) but reduced it slightly when nerve activity was abolished by tetrodotoxin (10 6 M) in the bottom panel., p 0.05;, p 0.01 compared with 100% (n 5). events do not use cgmp or protein kinase G (Ahern et al., 1999; Ignarro et al., 1999; Pinilla et al., 1999; Taglialatela et al., 1999; Garry et al., 2000; Janssen et al., 2000; Liu et al., 2000; Mazzuco et al., 2000; Takeda et al., 2000; Tseng et al., 2000). We aimed to evaluate the roles of guanylate cyclase in the downstream events initiated by NO released from nerves and from muscle. Materials and Methods Fig. 2. Effects if ODQ (10 5 M) added after TEA (20 mm) in the presence of nerve function (top panel) and in the absence of nerve function (bottom panel). TEA increased tone in the presence and in the absence of nerve function, in contrast to ODQ (Fig. 1). However, the three values obtained in the presence of active nerves were not quite significantly different when tested by one-way ANOVA (p 0.061), but the values obtained after block of nerve function were highly significantly different (p ). ODQ after TEA had no further effect with or without nerve function., p 0.05;, p 0.01 compared with controls (n 7). Tissue Preparation. Mongrel dogs of either sex were euthanized with an intravenous overdose of pentobarbital sodium (100 mg/kg), according to a protocol approved by the McMaster University Animal Care Committee and following the guidelines of the Canadian Council on Animal Care. The abdomen was opened along the midline, and segments of lower esophagus, ileum, and colon were excised and immediately put into oxygenated Krebs-Ringer solution at 24 C having the following composition: mm NaCl 2, 4.6 mm KCl, 1.2 mm MgSO 4, 22.0 mm NaHCO 3, 2.5 mm CaCl 2, and 11.0 mm glucose. The gastroesophageal junction was removed and opened along the greater curvature. After careful removal of the mucosa by fine dissection, the thickened ring of muscle, the LES, was removed. The mucosa was removed by fine dissection, leaving the muscularis externa. In Vitro Studies. In all cases circular muscle strips were prepared by cutting tissues into multiple 15 2 mm strips. These were tied with fine thread at both ends and mounted vertically in 5-ml organ baths, bathed in Krebs-Ringer solution at 37 C, and oxygenated with 95% O 2 and 5% CO 2. Strips were tied at the bottom to an electrode holder, passed through concentric platinum electrodes, and tied at the top to a force displacement transducer (Grass FT OC3; Grass Instruments, Quincy, MA). Tensions were recorded on Beckman R611 Dynagraphs (Beckman Coulter, Inc., Fullerton, CA). Electrodes were stimulated from a Grass 88 stimulator set at 40 V/cm, 5 pps, and 0.3-ms pulse duration, which gives near maximal relaxation of LES by activation of enteric nerves. LES strips had 2 g of tension applied and equilibrated for 1 h, during which the muscle strips contracted and spontaneously developed tone. Active tension was the difference between the observed tension and that obtained at the end of the experiment when Ca 2 - free Ringer s solution with 1 mm EGTA was applied. In all cases, we checked that the initial baseline set on the oscillograph remained unchanged by cutting the string from the tissue to the strain gauge. In some cases the deviation was more than 2 mm; because of this concern and possible Ca 2 -independent contractions, we used the level of tension obtained after this string was cut. Relaxation responses to electrical field stimulation (EFS) at 40 V/cm, 5 pps, and 0.1-, 0.2-, or 0.3-ms duration were executed until reproducible responses were obtained. Responses to 0.3 ms of EFS usually produced maximal relaxation and were used for statistical comparisons. Then 10 5 M ODQ (determined as maximal in preliminary experiments)

3 NO Mechanisms in Canine LES 1113 TABLE 1 Effects on tone Initial tone 100%. Means S.E.M. Drugs n Control Experiment p Value a Effect ODQ (10 5 M) * Effect ODQ after tetrodotoxin Drugs n Control TEA TEA ODQ p Value After TEA (20 mm) ODQ * * After TEA ODQ after tetrodotoxin (10 6 M) * * a p values were determined by paired comparisons except when three groups were considered; then one-way ANOVA with Dunnett s test was used. * Significantly different from 100%. or other agents were added at 20-min intervals, and effects on tone and on nadirs of EFS-induced relaxations were measured. If an agent lowered tone to the level of the nadir of relaxation, we added carbachol (10 6 M) to restore tone so that relaxation could be tested. At the end of each experiment, sodium nitroprusside at 10 4 M was added to evaluate the effects on NO from an exogenous donor. This Fig. 3. Representative tracings from experiments in which effects of ODQ were examined. All traces show baseline tone before applying 2 g of tension (not shown), tone development, responses to EFS, effects after 20 min on tone and response to EFS of drugs, if any were applied, then relaxation to sodium nitroprusside (10 4 M) after 20 min or more, and, finally, subsequent relaxation to calcium-free solution with 0.1 mm EGTA. A, control recording. Both the tone and the relaxation response to EFS was stable over time. Note that relaxation to sodium nitroprusside relaxed nearly to baseline (passive) tension. B, effect of ODQ (10 5 M). In the absence of 10 6 M tetrodotoxin, it increased tone and changed the response to EFS from mono- or biphasic relaxation to phasic contraction followed by a smaller slower relaxation and recovery. The relaxation to sodium nitroprusside was partially inhibited compared with control. C, effect of ODQ followed by L-NAME ( M). The effects of ODQ on tone, EFS, and sodium nitroprusside responses were not further changed after L-NAME. In this experiment, but not all, the relaxation to calcium-free EGTA solution was reduced (compare with B). D, effect of 20 mm TEA followed by ODQ. Note that TEA enhanced tone but did not inhibit relaxation to EFS. Previous studies (Daniel et al., 2000) showed that it reduced relaxation to sodium nitroprusside about 50%. Subsequent ODQ now abolished all responses to EFS and nearly abolished responses to sodium nitroprusside. Subsequent relaxation to calcium-free solution with EGTA solution was complete.

4 1114 Daniel et al. TABLE 2 Relaxation nadir to EFS Initial tone 100%. Means S.E.M. Drugs n Control Experiment p Value a After ODQ (10 5 M) After ODQ After ODQ After L-NOARG (10 4 M) vs. after ODQ After L-NOARG (10 4 M) vs. after ODQ Drugs n Control TEA (20 mm) TEA ODQ p Value After TEA ODQ after tetrodotoxin (10 6 M) a p values were determined by paired comparisons except when three groups were considered; then one-way ANOVA with Dunnett s test was used. Significantly different from 0 and 100%. was followed by Ca 2 -free Ringer s solution with 1 mm EGTA to eliminate active tone. A typical protocol was as follows: Tension 3 Tone Development 3 EFS 3 Experimental Agent 3 EFS 3 Sodium Nitroprusside 3 EGTA 3 Cut String If tetrodotoxin was used, EFS was retested afterward and from time to time during experiments. Patch-Clamp Techniques. The LES was dissected as described above and strips were cut into 1- to 2-mm 2 square pieces and placed in the dissociation solution. Cell Isolation. Cells were dissociated in 0.25 mm EDTA, 125 mm NaCl, 4.8 mm KCl, 1 mm CaCl 2, 1 mm MgCl 2, 10 mm HEPES, and 10 mm glucose for 30 min. An enzyme solution containing papain (130 mg ml 1 ), 1,4-dithio-L-threitol, (15.4 mg ml 1 ), bovine serum albumin (100 mg ml 1 ), and Sigma collagenase blend H (occasionally F) was added to the tissue pieces for 30 to 60 min. After incubation, the enzyme solution was decanted off, and the tissue pieces were rinsed in enzymefree dissociation solution. Single cells were gently mechanically agitated with siliconized Pasteur pipettes to disperse and isolate single smooth muscle cells. Cells used in this study were patch clamped at room temperature (22 24 C) usually within 8 h of isolation. Patch-Clamp Methodology. Cells from the suspension were placed in a glass-bottomed dish. Within 30 min, cells adhered to the dish. The cells were then washed by perfusion with Ca 2 containing external solution that contained 140 mm NaCl, 4.5 mm KCl, 2.5 mm CaCl 2, 1 mm MgCl 2, 10 mm HEPES, and 5.5 mm glucose, ph adjusted to 7.35 with NaOH. Patch electrodes were made using borosilicate glass capillary tubes and a Flaming Brown micropipette puller (Sutter Instrument Co., Novato, CA). After polishing with a microforge (Narishige MF-830; W. Nuhsbaum, Inc., McHenry, IL) and filling, pipettes had resistances of 2 to 5 M. High Ca 2 pipette solution contained 2.5 mm CaCl 2, 140 mm KCl, 1 mm MgCl 2, 10 mm HEPES, 4 mm sodium-atp, and 0.3 mm EGTA to obtain free Ca 2 of 8 M. CaCl 2, KCl, and EGTA levels were adjusted to obtain 200 nm free Ca 2 levels as calculated using MAX Chelator software (version 6.72) by Bers et al. (1994). A standardized stimulation protocol was used to evoke currents from isolated smooth muscle cells, which were studied without leak subtraction. All cells had access resistance 25 m. Cell capacitances averaged 58 5pF(n 6) for canine LES cells. Cells were held at 50 mv and subsequently depolarized in seven cumulative steps, each 250 ms in duration, of 20 mv. Current/voltage curves were constructed using the maximum current values measured at t 200 ms in the pulse. Membrane currents were measured with an Axopatch 1C voltage clamp amplifier (Axon Instruments, Union City, CA), filtered with a 0.3 db Bessel filter at 1 khz, and recorded on-line using pclamp 5.5b software. Drugs Used. ODQ was obtained from Sigma-Aldrich Canada (Oakville, ON, Canada). All other drugs were purchased from Sigma- Aldrich (St. Louis, MO). Data Analysis. Tone and relaxation nadirs were measured in LES in each strip. Each was compared with the control values as 100% and to the observed initial nadir of relaxation. Changes with concentrations of inhibitor were evaluated using Dunnett s multiple comparisons test unless only one concentration was used. Then paired comparisons were made. All statistical tests were carried out using Prism 3 software (Intersil Corp., Irvine, CA). The n values in tables refer to the number of animals that supplied the tissues. Fig. 4. Comparison of the effects of ODQ (10 5 M) and L-NOARG ( M) on the nadir of relaxation to EFS. Both raised the nadir, but L-NOARG raised it to values not different from 100% (no relaxation), whereas ODQ raised it significantly but not to 100%., p 0.01, compared with controls (n 5, 4). Results Drug Effects on Active Tone with/without Functioning Nerves. In the absence of nerve blockade with 10 6 M tetrodotoxin, ODQ (10 5 M) and TEA (20 mm) both increased active tone, TEA more than ODQ (cf. Figs. 1, 2, and 3, B and C). Tetrodotoxin alone or combined with -conotoxin (GVIA) had no effect on tone (Salapatek et al., 1998b). When tetrodotoxin had blocked responses to electrical field stimulation, TEA, but not ODQ, still increased active tone (Fig. 2 and

5 NO Mechanisms in Canine LES 1115 Fig. 5. Relaxation after 20 mm TEA and after subsequent ODQ (10 5 M). Results expressed in terms of the magnitudes of relaxation instead of the nadirs of relaxation (Table 2). After TEA, which increased tone, relaxation was increased, but following the addition of ODQ, it was abolished (not significantly different from 0). One-way ANOVA gave p (n 7). Fig. 6. Comparison of the effects of ODQ and L-NOARG on nadirs of relaxation to sodium nitroprusside (10 4 M) and to calcium-free EGTA (0.1 mm). Previous studies showed that L-NOARG has no effect on sodium nitroprusside-induced relaxation (Daniel et al., 2000). However, as shown in the top panel, ODQ increased the nadir of relaxation to sodium nitroprusside significantly, compared with L-NOARG. It did not significantly affect the nadir to removal of extracellular calcium as shown in the bottom panel., p 0.05 unpaired comparison (n 6, 5). Table 1), and ODQ after TEA had no additional effect. Moreover, the increase in tone from TEA was greater after tetrodotoxin than before, presumably a result of removal of TEAinduced release of inhibitory nerve mediator. Drug Effects on Relaxation to EFS. ODQ (10 5 M) markedly reduced relaxation to EFS (40 V/cm, 5 pps, 0.3-ms duration), converting the relatively rapid relaxation to a phasic contraction followed by a small, slow relaxation (Fig. 3, B and C) not further affected by L-NAME (Fig. 3C). The inhibition of relaxation is summarized in Table 2. Figure 4 shows that ODQ, in contrast to L-NOARG ( M), reduced but did not abolish relaxation to EFS. In contrast to ODQ and confirming previous findings (Daniel et al., 2000), 20 mm TEA had no effect to inhibit the amplitude of relaxation when present alone (Fig. 5), although it did reduce the extent of relaxation in some experimental series (Table 2). However, when TEA was present together with ODQ, relaxation to EFS was abolished (Figs. 3D and 5), as it was with L-NOARG. Drug Effects on Relaxation to Sodium Nitroprusside and Calcium-Free (EGTA) Solution. L-NOARG had no effect (Fig. 6 and Table 3) on relaxation to sodium nitroprusside ( M), irrespective of the presence or absence of tetrodotoxin (not shown). ODQ or TEA each reduced relaxation to sodium nitroprusside, and together they markedly reduced it but did not abolish it (Figs. 3D and 7), unrelated to the presence or absence of tetrodotoxin (not shown). In most cases, the residual tone (assumed to be passive) after EGTA was the same irrespective of any pretreatment (Fig. 6 and Table 4). However, when tetrodotoxin was present, the combination of ODQ and TEA significantly increased it (Fig. 7). Patch-Clamp Study. ODQ (10 5 M) had no effect on outward currents when the patch pipette had 200 nm free Ca 2 (Fig. 8) This is in contrast to effects of TEA (20 mm) or L-NOARG ( M), shown previously (Salapatek et al., 1998c) and confirmed here (not shown) to inhibit them by

6 1116 Daniel et al. TABLE 3 Tone nadir after sodium nitroprusside (10 4 M) Initial tone 100%. Means S.E.M. Experiment n Control Experiment p Value a L-NOARG (10 4 M) 6, TEA/ODQ vs. 20 mm/10 5 M TEA/ODQ/tetrodotoxin (10 6 M)** a p values were determined by paired comparisons. Significantly different from 0 and 100%. Both were significantly different from tone before sodium nitroprusside, not from 100%. ** TEA alone reduced relaxation to sodium nitroprusside to a nadir of 50%. about 80% at maximum depolarization. We also reconfirmed that sodium nitroprusside (10 4 M) restored the currents inhibited by L-NOARG (Salapatek et al., 1998c). Fig. 7. Nadirs of relaxation to sodium nitroprusside (top panel) or calcium-free EGTA (0.1 mm) (bottom panel) after TEA and ODQ irrespective of the presence (right histogram) or absence (left histogram) of nerve activity. Although the final tone levels achieved after sodium nitroprusside were not significantly different from 100% initial tone in the presence or absence of nerve activity, there was a small relaxation since TEA had increased tone from the initial level (Fig. 2). The bottom panel shows that relaxation to calcium-free solution was also impaired significantly by TEA with ODQ when tetrodotoxin was present., p 0.01 paired comparison (n 7). TABLE 4 Tone nadir to Ca 2 -free (EGTA) solution Initial tone 100%. Means S.E.M. Discussion ODQ is a widely used inhibitor of cytosolic guanylate cyclase (Ignarro et al., 1999; Chang et al., 2000; Gorodeski, 2000a,b; Janssen et al., 2000; Kwan et al., 2000; Tseng et al., 2000; Yao et al., 2000). In this study 10 5 M ODQ increased tone and raised the nadir of NO-mediated relaxation (Jury et al., 1992) to nerve stimulation (EFS) from 38.5 to 77.5% of initial tone and changed the response qualitatively. Instead of a large fast relaxation, there was a small initial phasic contraction followed by a slow, small relaxation (Fig. 3, B and C). This relaxation was unaffected by subsequent L-NAME, suggesting that it is not NO-mediated. After tetrodotoxin abolished responses to EFS, the tone increase to ODQ and the altered EFS response following ODQ were also eliminated, suggesting that both were nervemediated. Reduced responses to basal release of neural NO, acting either on interstitial cells of Cajal (ICC) or on nerve endings, likely accounts for the tone increase after ODQ (see discussion below). Previous studies (Allescher et al., 1988) showed that stimulation of intrinsic nerves of LES releases acetylcholine as well as NO. Acetylcholine was the probable mediator of contraction, observed when actions of neural NO were reduced by ODQ. It also may have contributed to the increase in tone when ODQ was administered with nerves active. Although not mediated by NO, the residual relaxation after ODQ involves opening of K channels, since it was abolished when TEA was present together with ODQ. We did not evaluate the efficacy of ODQ to block elevations of cgmp levels, so it may have inhibited some responses nonselectively, or some tissue components may have resisted its effects. However, it selectively inhibited two NO-mediated responses (relaxation to EFS and to sodium nitroprusside). Thus, it is unlikely that the failure to inhibit NO actions when NO was derived from muscle can be explained by resistance to ODQ. It did increase spontaneous tone when nerves were functioning, but we attribute this effect to inhibition of the action of NO released from nerves under basal conditions because it disappeared after tetrodotoxin. Thus, Experiment n Control Experiment p Value a L-NOARG (10 4 M) vs. ODQ (10 5 M) 6, TEA/ODQ vs. (20 mm/10 5 M) TEA/ODQ (tetrodotoxin) a p values were determined by paired comparisons. Values for other groups are not shown; none were significantly different from values in left (control) column. Significantly different from 0.

7 NO Mechanisms in Canine LES 1117 Fig. 8. Current/voltage plots of the lack of effect of 10 5 M ODQ on outward currents induced by depolarizing steps (20 mv, 250 ms) from 50 to 90 mv. Either L-NOARG or TEA reduced these currents by 70 to 80% (not shown), confirming previous results (Salapatek et al., 1998c). its actions appeared efficacious against some NO-mediated events and selective in that it failed to act on other responses. We postulate that ODQ eliminates most or all of the effects of NO released from nerves, disinhibiting (Fox-Threlkeld et al., 1999) and/or unmasking the response to the release of acetylcholine previously obscured by NO actions. We speculate that nerves release two inhibitory mediators, one NO and the other released in response to NO release. Dependence on NO release would explain why L-NAME alone usually completely inhibits relaxation to low-frequency (1 5 pulses per second) EFS (Jury et al., 1992). According to this model, neural NO acts on cgc to increase cgmp levels, but the other mediator acts to open K channels. The combination of ODQ and TEA would, on either model, inhibit both the NO/cGC and the K channel-mediated components of the response. NO released from nerves may act on cgc in muscle or on the ICC. ICC are present in canine LES, intercalated between nerve endings and muscle to which they are connected by gap junctions (Allescher et al., 1988). Similar structural relations are found in other species and parts of the gastrointestinal tract (e.g., Daniel and Posey-Daniel, 1984; Ward et al., 1998; reviewed in Sanders et al., 1999). Recent studies in mutant W/W V mice with apparently normal nitric-oxide synthase innervation of the LES, but lacking intramuscular ICC (Ward et al., 1998), found that nerve stimulation did not induce the usual relaxation and hyperpolarization from NO release and that NO donors also became less effective to induce relaxation and hyperpolarization in the absence of ICC. This dependence of NO-mediated neural effects on ICC was also found in the pylorus and gastric fundus (Burns et al., 1996; Ward et al., 1998). Thus, NO released from nerves may activate cgc in ICC rather than in smooth muscle. The NO-mediated activation of cgc may produce hyperpolarization in ICC cells, which is transmitted to LES muscle by gap junctions or by other means. Studies of responses of intramuscular ICC to NO are lacking because of the technical difficulty of isolating them. However, inhibitors of gap-junction conductance did not interfere with nerve-mediated relaxation of the LES (Daniel et al., 2001b). LES gap junctions had connexins 43 and 40 colocalized in them (Wang and Daniel, 2001), and the conductance properties of such junctions are unknown. Ionic mechanisms involved in neural NO effects are incompletely understood. Previous studies showed that NO-mediated hyperpolarization accompanied relaxation of muscle and appeared to be associated with current flow through K channels (Christinck et al., 1991; Jury et al., 1992; Cayabyab and Daniel, 1995). Although 20 mm TEA alone had no inhibitory effect on amplitudes of nerve-mediated relaxation, it reduced outward currents 80% in LES cells (Salapatek et al., 1998c). Whether LES relaxation after TEA was still associated with hyperpolarization has not been determined. The sites of K channels, ICC or muscle, affected by NO from nerves or the secondary inhibitory mediator are unclear. Clearly, the mechanisms by which LES relaxation and hyperpolarization are induced after activation of intrinsic nerves are complex and need further study. Previous studies showed that activation of BK Ca channels driven by release of NO from a membrane-bound nitric-oxide synthase accounted for large outward currents, cellular hyperpolarization, and reduced tone in canine LES (Salapatek et al., 1998b,c). These currents were inhibited 80% by L- NOARG, iberiotoxin, or TEA. They were restored by sodium nitroprusside after L-NOARG but not after the K channel blockers. The nitric-oxide synthase activity depended upon Ca 2 entering through L-type calcium channels, closely associated with the nitric-oxide synthase in caveolae (Daniel et al., 2001a). The tone modulation from nitric-oxide synthase activity in LES strips was abolished by L-NOARG and 20 mm TEA but only partially by iberiotoxin (Daniel et al., 2000). This suggested that additional K channels besides BK Ca were activated by NO of myogenic origin. ODQ had no effect on tone in the presence of tetrodotoxin and no effect on outward currents. Thus, there appears to be no role for cgc in the modulation of tone and enhancement of outward currents by NO from myogenic nitric-oxide synthase. We did not test that cgc was still active in isolated LES cells. However, the actions of NO to drive outward currents do not require its activity. L-NOARG reduces the outward currents by 80% and NO donors restore them. NO in other systems also operates independently of cgc (Ahern et al., 1999; Ignarro et al., 1999; Pinilla et al., 1999; Taglialatela et al., 1999; Garry et al., 2000; Janssen et al., 2000; Liu et al., 2000; Mazzuco et al., 2000; Takeda et al., 2000; Tseng et al., 2000).

8 1118 Daniel et al. NO from the muscle membrane may act locally on nearby ion channels; examples of such local actions have been reported (Bolotina et al., 1994). The NO-dependent outward currents depend on Ca 2 entry through L-type calcium channels (Salapatek et al., 1998b,c). However, in opossum LES Akbarali and Goyal (1994) found that NO inhibits L-type calcium currents. Nifedipine abolishes tone in canine LES. However, inward currents opened by depolarization and carried by L-type calcium channels could not be demonstrated in LES cells from dog or opossum, even after all potassium currents were blocked (Jury and Daniel, 1996; J. Jury and E. E. Daniel, unpublished data), but they were easily demonstrated in the body of opossum esophagus. Yet, as noted above, NO-driven BK Ca currents in LES are inhibited when L-type calcium channels are blocked. The presence of these channels in close proximity to nitric-oxide synthase in caveolae (Daniel et al., 2001a) may result in unexpected properties, including altered voltage dependence and opening properties. Relaxation to sodium nitroprusside, a source of exogenous NO, was reduced by TEA or by ODQ. Together, they nearly eliminated it. This implies that exogenous NO acts by both activating cgc and opening K channels independently of cgc. Since ODQ had no effects on NO-driven outward currents in LES cells, TEA likely inhibited exogenous NO actions at sites on which myogenic-derived NO acts, and ODQ inhibited them on sites at which NO from nerves acts, possibly in ICC. Exogenous NO had access to both sites. In a previous study (Daniel et al., 2000) we found that TEA induced a tone increase when nerve function was abolished and occluded the tone increase to L-NOARG, consistent with an action of TEA to block effects of myogenic NO release. However, cgc was not demonstrated to be present or responsive to NO in isolated LES cells. If it is absent or nonresponsive to NO in LES cells in situ but present in ICC, this would provide an explanation for the differential effects of ODQ in vitro noted above. In summary, these findings show that relaxation of the canine LES by neural NO is more complex than originally presumed; i.e., NO from nerves appears to activate cgc, likely in ICC, whereas NO from muscle apparently does not act through cgc. NO from muscle activates K channels, whereas NO from nerves causes relaxation, which does not require opening of TEA-sensitive potassium channels. NO added exogenously acts through both ODQ- and TEA-sensitive mechanisms. References Ahern GP, Hsu SF, and Jackson MB (1999) Direct actions of nitric oxide on rat neurohypophysial K channels. J Physiol (Lond) 520: Akbarali HI and Goyal RK (1994) Effect of sodium nitroprusside on Ca2 currents in opossum esophageal circular muscle cells. Am J Physiol 266:G1036 G1042. Allescher H-D, Berezin I, Jury J, and Daniel EE (1988) Characteristics of canine lower esophageal sphincter: a new electrophysiological tool. Am J Physiol 255: G441 G453. Bazan-Perking B, Carbajal V, Sommer B, Gonzalez-Martinez M, Valenzuela F, Daniel EE, and Montano LM (1998) Involvement of different Ca 2 pools during the canine bronchial sustained contraction in Ca 2 free medium: lack of effect of PKC inhibition. Naunyn-Schmiedeberg s Arch Pharmacol 358: Bers DM, Patton CW, and Nuccitelli R (1994) A practical guide to the preparation of Ca2 buffers. Methods Cell Biol 40:3 29. Bolotina VM, Najibi S, Palacino JJ, Pagano PJ, and Cohen RA (1994) Nitric oxide directly activates calcium-dependent potassium channels in vascular smooth muscle. Nature (Lond) 368: Burns AJ, Lomax AE, Torihashi S, Sanders K, and Ward SM (1996) Interstitial cells of Cajal mediate inhibitory neurotransmission in the stomach. Proc Natl Acad Sci USA 93: Cayabyab FS and Daniel EE (1995) K channel opening mediates hyperpolarizations by nitric oxide donors and IJPs in opossum esophagus. Am J Physiol 268: G831 G842. Chang YS, Yaccino JA, Lakshminarayanan S, Frangos JA, and Tarbell JM (2000) Shear-induced increase in hydraulic conductivity in endothelial cells is mediated by a nitric oxide-dependent mechanism. Arterioscler Thromb Vasc Biol 20: Christinck F, Jury J, Cayabyab F, and Daniel EE (1991) Nitric oxide may be the final mediator of non-adrenergic non-cholinergic inhibitory junction potentials in gut. Can J Physiol Pharmacol 69: Daniel EE, Jury J, Salapatek AM, Bowes T, Lam A, Thomas S, Ramnarain M, Nguyen V, and Mistry V (2000) Nitric oxide from enteric nerves acts by a different mechanism from myogenic nitric oxide in canine lower esophageal sphincter. J Pharmacol Exp Ther 294: Daniel EE, Jury J, and Wang YF (2001a) nnos in canine lower esophageal sphincter: colocalized with Cav-1 and Ca 2 -handling proteins. Am J Physiol Gastrointest Liver Physiol 281:G1101 G1114. Daniel EE and Posey-Daniel V (1984) Neuromuscular structures in opossum esophagus: role of interstitial cells of Cajal. Am J Physiol 246:G305 G315. Daniel EE, Thomas J, Ramnarain M, Bowes TJ, and Jury J (2001b) Do gap junctions couple interstitial cells of Cajal pacing and neurotransmission to gastrointestinal smooth muscle. Neurogastroenterol Motil 13: Darby PJ, Kwan CY, and Daniel EE (2000) Caveolae from canine airway smooth muscle contain the necessary components for a role in Ca 2 handling. Am J Physiol Lung Cell Mol Physiol 279:L1226 L1235. Fox-Threlkeld JET, McDonald TJ, Woskowska Z, and Daniel EE (1999) Pituitary adenylate cyclase-activating peptide as a neurotransmitter in the canine ileal circular muscle. J Pharmacol Exp Ther 290: Garry MG, Walton LP, and Davis MA (2000) Capsaicin-evoked release of immunoreactive calcitonin gene-related peptide from the spinal cord is mediated by nitric oxide but not by cyclic GMP. Brain Res 861: Gorodeski GI (2000a) Role of nitric oxide and cyclic guanosine 3,5 -monophosphate in the estrogen regulation of cervical epithelial permeability. Endocrinology 141: Gorodeski GI (2000b) NO increases permeability of cultured human cervical epithelia by cgmp-mediated increase in G-actin. Am J Physiol Cell Physiol 278:C942 C952. Ignarro LJ, Cirino G, Casini A, and Napoli C (1999) Nitric oxide as a signaling molecule in the vascular system: an overview. J Cardiovasc Pharmacol 34: Janssen LJ, Premji M, Lu-Chao H, Cox G, and Keshavjee S (2000) NO( ) but not NO radical relaxes airway smooth muscle via cgmp-independent release of internal Ca(2 ). Am J Physiol Lung Cell Mol Physiol 278:L899 L905. Jury J, Ahmedzadeh N, and Daniel EE (1992) A mediator derived from arginine is released from sphincteric intrinsic nerves to mediate inhibitory junction potentials and relaxations. Can J Physiol Pharmacol 70: Jury J and Daniel EE (1996) A comparison between isolated smooth muscle cells of the opossum esophageal body (BCM) and lower esophageal sphincter (LES) (Abstract). Gastroenterology 110:A689. Kwan HY, Huang Y, and Yao X (2000) Store-operated calcium entry in vascular endothelial cells is inhibited by cgmp via a protein kinase G-dependent mechanism. J Biol Chem 275: Liu XD, Skold CM, Umino T, Spurzem JR, Romberger DJ, and Rennard SI (2000) Sodium nitroprusside augments human lung fibroblast collagen gel contraction independently of NO-cGMP pathway. Am J Physiol Lung Cell Mol Physiol 278: L1032 L1038. Mazzuco TL, Andre E, and Calixto JB (2000) Contribution of nitric oxide, prostanoids and Ca(2 )-activated K channels to the relaxant response of bradykinin in the guinea pig bronchus in vitro. Naunyn-Schmiedeberg s Arch Pharmacol 361: Montano LM, Barajas-Lopez C, and Daniel EE (1996) Canine bronchial sustained contraction in Ca 2 free medium: role of cartilage as a source of Ca 2 and recycling of intracellular Ca 2. Can J Physiol Pharmacol 74: Montano LM, Jones GL, O Byrne PM, and Daniel EE (1993) The effect of ozone exposure in vivo on response of bronchial rings in vitro; the role of intracellular Ca 2. J Appl Physiol 75: Pinilla L, Tena-Sempere M, and Aguilar E (1999) Nitric oxide stimulates growth hormone secretion in vitro through a calcium- and cyclic guanosine monophosphate-independent mechanism. Horm Res 51: Salapatek AM, Lam A, and Daniel EE (1998a) Calcium source diversity in canine lower esophageal sphincter muscle. J Pharmacol Exp Ther 287: Salapatek AM, Wang YF, Mao YK, Lam A, and Daniel EE (1998b) Myogenic nitric oxide synthase activity in canine lower oesophageal sphincter: morphological and functional evidence. Br J Pharmacol 123: Salapatek AMF, Wang YF, Mao YK, Mori M, and Daniel EE (1998c) Myogenic NOS in canine lower esophageal sphincter: enzyme activation, substrate recycling and product actions. Am J Physiol Cell Physiol 274:C1145 C1157. Sanders KM, Ordog T, Koh SD, Torihashi S, and Ward SM (1999) Development and plasticity of interstitial cells of Cajal. Neurogastroenterol Motil 11: Taglialatela M, Pannaccione A, Iossa S, Castaldo P, and Annunziato L (1999) Modulation of the K( ) channels encoded by the human ether-a-gogo-related gene-1 (herg1) by nitric oxide. Mol Pharmacol 56: Takeda K, Asayama J, Fliss H, and Nakagawa M (2000) Cytokine-induced nitric oxide production inhibits mitochondrial energy production and impairs contractile function in rat cardiac myocytes. J Am Coll Cardiol 35: Tseng CM, Tabrizi-Fard MA, and Fung HL (2000) Differential sensitivity among nitric oxide donors toward ODQ-mediated inhibition of vascular relaxation. J Pharmacol Exp Ther 292: Wang YF and Daniel EE (2001) Gap junctions in gastrointestinal muscle contain multiple connexins. Am J Physiol Gastrointest Liver Physiol 281:G533 G543. Ward SM, Morris G, Reese L, Wang XY, and Sanders KM (1998) Interstitial cells of Cajal mediate enteric inhibitory neurotransmission in the lower esophageal and pyloric sphincters. Gastroenterology 115: Yao X, Kwan HY, Chan FL, Chan NW, and Huang Y (2000) A protein kinase G-sensitive channel mediates flow-induced Ca(2 ) entry into vascular endothelial cells. FASEB J 14: Address correspondence to: Dr. E. E. Daniel, Department of Pharmacology, University of Alberta, 9-70 Medical Sciences Building, Edmonton, AB T6G 2H7, Canada. edaniel@ualberta.ca