Nitric oxide generates slow electrical oscillations (SEOs) in cells near the myenteric edge of the circular muscle layer, which resemble slow waves generated by interstitial cells of Cajal (ICCs) at the submucosal edge of this muscle. The properties of SEOs were studied to determine whether these events are similar to slow waves. Rapid frequency membrane potential oscillations (MPOs; 16 ± 1 cycles/min and 9.6 ± 0.2 mV) were recorded from control muscles near the myenteric edge. Sodium nitroprusside (0.3 μM) reduced MPOs and initiated SEOs (1.3 ± 0.3 cycles/min and 13.4 ± 1.4 mV amplitude). SEOs were abolished by the guanylate cyclase inhibitor 1H-[1,2,4]-oxadiazolo-[4,3-a]-quinoxaline-1-one (10 μM). MPOs were abolished by nifedipine (1 μM), whereas SEO frequency increased and the amount of depolarization decreased. BAY K 8644 (1 μM) prolonged SEOs and reduced their frequency. SEOs were abolished by Ni2+ (0.5 mM), low Ca2+ solution (0.1 mM Ca2+), cyclopiazonic acid (10 μM), and the mitochondrial uncouplers antimycin (10 μM) and carbonyl cyanidep-trifluoromethoxyphenylhydrazone (1 μM). Oligomycin (10 μM) was without effect. These effects are similar to those described for colonic slow waves. Our results suggest that nitric oxide-induced SEOs are similar in mechanism to slow waves, an activity not previously thought to be generated by myenteric pacemakers.
- smooth muscle
- interstitial cells of Cajal
- slow wave
- colonic motility
- enteric nervous system
electrical slow waves arise from interstitial cells of Cajal (ICCs) at the submucosal edge of the circular muscle layer in the canine colon (11, 12,22). These events occur at a regular frequency of 4–6/min. Slow waves play an important role in determining motility patterns in the colon. Recent studies have suggested that the inward current responsible for slow wave depolarization is due to nonselective cation channels expressed by ICCs (5, 14). Periodic activation of the cation channels is due to intracellular Ca2+ release from the endoplasmic reticulum and uptake of Ca2+ by mitochondria (18).
A second electrical rhythm is generated near the surface of the circular muscle layer adjacent to the myenteric plexus, and therefore it has been suggested that two electrical pacemakers contribute to the electrical behavior of the colonic tunica muscularis (11). The myenteric pacemaker generates a fast electrical rhythm referred to as membrane potential oscillations (MPOs), and these events occur at 17 cycles/min (11). A second plexus of ICCs is present at the myenteric edge of the circular muscle layer (1,22), and removing a thin strip of muscle containing the myenteric ICCs abolishes MPO activity (11). The myenteric network of ICCs may be the source of pacemaker activity in this region of the muscularis, but this hypothesis has not been rigorously tested. Previous studies have implied that there are significant differences in the properties and, possibly, the mechanisms responsible for the two colonic pacemakers.
MPO activity is not stable, and we previously showed that release of nitric oxide (NO) from intrinsic nerves (or the addition of exogenous NO) can switch the pattern of fast MPOs into a repetitive cycle of slow electrical oscillations [SEOs (3)]. Although quite different from MPOs, these SEOs qualitatively resemble the slow waves recorded from cells near the submucosal surface of the circular muscle. This observation suggests that there may be fundamental similarities between the two colonic pacemakers at the myenteric and submucosal surfaces of the circular muscle that are not immediately apparent in unstimulated muscles. The fact that stimulation of intrinsic nerves can switch MPO activity into an alternate electrical pattern makes it important to understand this process. The purpose of the present study was to examine the nature of SEOs and to determine whether the mechanisms giving rise to SEOs are similar to those that generate slow waves in the submucosal region.
Male and female mongrel dogs were killed with an overdose of pentobarbital sodium (100 mg/kg). This method is consistent with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association and was approved by the Institutional Animal Care and Use Committee. The abdomen was opened, and a segment of proximal colon 6–14 cm from the ileocecal sphincter was removed. The colon was opened along the mesenteric border, cleared of remaining fecal material, and pinned out in a dissecting dish containing oxygenated Krebs-Ringer bicarbonate solution (KRB) of the following composition (in mM): 118.5 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgCl2, 23.8 NaHCO3, 11.0 KH2PO4, and dextrose. This solution had a pH of 7.4 at 37°C when bubbled to equilibrium with 95% O2-5% CO2. Unless otherwise specified, experiments were performed in the presence of 100 μMN ω-nitro-l-arginine (l-NNA).
Strips (15 mm) of the tunica muscularis were cut parallel to the circular muscle fibers with a knife consisting of a pair of parallel scalpel blades set 1.5 mm apart. The mucosa was removed from strips by sharp dissection. After dissection, the muscle strips were pinned in cross-section to the floor of an electrophysiological chamber. The strip was then cut in two halves, and the half that contained the myenteric plexus and longitudinal muscle was retained; the submucosal half was discarded. Muscle strips were superfused with warmed, oxygenated KRB solution. Temperature was monitored by a thermistor probe submerged in the perfusion solution near the muscle and was maintained at 37 ± 0.5°C. Muscles were allowed to equilibrate for ∼2 h before experiments were undertaken.
Muscle cells were impaled with glass microelectrodes filled with 3 M KCl and having resistances ranging from 60 to 100 MΩ. Membrane potential was measured with a high input impedance electrometer (model Duo 773; World Precision Instruments, New Haven, CT), and outputs were displayed on an oscilloscope (Hitashi). Analog electrical signals were digitized and recorded on videotape (Vetter 875) and AcqKnowledge 3.2.4 software (Biopac System) for later data analysis.
Analysis of data.
Several different parameters of electrical activity were tabulated. Resting membrane potential of cells in the myenteric region was determined as the average of the most negative values occurring between MPOs. Peak depolarization of MPOs was determined as the average maximal level of depolarization reached during 10 consecutive MPOs. “Resting” membrane potential of muscles during SEOs was defined as the average of the most negative values of membrane potential occurring between oscillations. The amplitude of slow oscillations was determined by subtracting resting membrane potential from the average of the most negative membrane potentials attained during the depolarization phase of slow oscillation. Because the SEO waveform was often complex, we quantitated changes in SEO time course by determining the entire amount of depolarization that occurred during the SEO. This was accomplished by integrating the voltage signal as a function of time between 10% depolarization and 90% repolarization.
Significant differences between means were calculated by a two-tailed paired or unpaired Student t-test, and values were considered significantly different when P < 0.05. Only one muscle strip per animal was used. Thus n values represent both the number of animals and the number of muscle strips tested.
Drugs used in this study included tetrodotoxin, l-NNA, sodium nitroprusside (SNP), antimycin, oligomycin, carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP), nifedipine, BAY K 8644 (all purchased from Sigma, St. Louis, MO), and 1H-[1,2,4]-oxadiazolo-[4,3-a]-quinoxaline-1-one (ODQ; Tockris Cookson, St. Louis, MO).
Effect of NO on electrical activity in myenteric muscle strips.
As we previously showed under controlled conditions, 25% of colonic muscle strips exhibited electrical events termed SEOs. These events were abolished by the neural blocker tetrodotoxin (TTX, 1 μM) and by NO synthase inhibitors (e.g., l-NNA, 100 μM), suggesting that this activity arises from spontaneous release of NO from intrinsic neurons (3). Examples of SEOs are shown in Fig.1, A and B. In the remainder of muscles (75%), the predominant activity was rapid MPOs with an average frequency of 16 ± 1 cycles/min and an average amplitude of 9.6 ± 0.2 mV, as previously documented (11). SEOs could be evoked in the majority of muscles when the tissues were stimulated with SNP (0.3–1 μM). SEOs were characterized by a sharp upstroke depolarization and a sustained, plateau-like depolarization. In many cases, oscillatory changes in potential were superimposed on the plateau phase of the SEO. The frequency of SEOs averaged 1.3 ± 0.3 cycles/min and amplitude averaged 13.4 ± 1.4 mV. The maximum level of polarization between SEOs averaged −65 ± 1.3 mV (n = 22). SEOs evoked by SNP were identical in waveform and characteristics to the SEOs that occurred spontaneously in 25% of muscles, which were previously shown to be due to neural release of NO [see Ref. 3 and Fig.1 C]. Thus because the purpose of this study was to characterize the nature of SEOs, we evoked this type of electrical behavior with SNP in all experiments, unless otherwise indicated. Muscles were also pretreated with l-NNA (100 μM) to reduce variations in NO levels as a function of endogenous production. SEOs had waveform parameters reminiscent of slow waves generated at the submucosal surface of the circular muscle layer (12). Therefore, we compared the effects of several agents known to affect slow waves to determine whether SEOs are due to a similar mechanism.
The block of SEOs by both TTX and l-NNA was associated with depolarization of resting membrane potential (Fig. 1, A andB). Furthermore, other experimental protocols outlined below also produced some membrane depolarization. To test whether depolarization per se leads to block of SEOs, we examined the effects of raising extracellular K+ concentration ([K]o). Increasing [K]o from 5.9 to 16.9 mM led to a 18.5 ± 0.6 mV (n = 4) depolarization of resting membrane potential. Although SEOs were reduced to 39 ± 7.5% of the control amplitude (n = 4), they were not abolished (Fig. 2 A). To further investigate the role of membrane potential in the generation of SEOs, additional experiments were undertaken in which preparations were hyperpolarized by the addition of lemakalim (Lem) instead of SNP. Lem activates ATP-sensitive K+ channels (21). Lem (3 μM) caused a 13 mV of hyperpolarization (from −56.9 ± 3.4 to −69.8 ± 1.6 mV) but did not elicit SEOs (n = 5, Fig. 2 B).
SEOs induced by NO were due to a mechanism involving guanylyl cyclase, because treatment of muscles with ODQ (10 μM) completely inhibited the ability of SNP to produce SEOs. Control-like activity was re-established in muscles treated with ODQ in the presence of SNP (n = 4; Fig. 3). In contrast, submucosal slow waves have previously been reported to be insensitive to ODQ (2).
Effects of dihydropyridines on SEOs.
Slow waves in colonic muscles are not blocked by dihydropyridines (19). Recent studies have reinforced this notion by showing that inward currents in pacemaker ICCs are due to nifedipine-insensitive cation channels (5, 14, 20). Therefore, we tested the effects of nifedipine (1 μM) on MPOs (observed typically in control muscles) and SEOs evoked by SNP. Nifedipine completely blocked MPOs (n = 4) and depolarized membrane potential from −56.5 ± 2.2 to −51.5 ± 2.8 mV (P < 0.05; Fig.4 A). In contrast, SEOs persisted in the presence of nifedipine (Fig. 4 B) although the amount of depolarization during the SEOs decreased from 0.3 ± 0.05 to 0.15 ± 0.04 mV/s (P < 0.05), membrane potential depolarized from −63.8 to −61.9 mV (P < 0.05), and SEO frequency increased from 1.3 ± 0.1 to 1.7 ± 0.4 cycles/min (P < 0.05; n = 9). Rapid oscillations in membrane potential superimposed on some SEOs under control conditions were abolished in the presence of nifedipine.
We also tested the effects of the L-type Ca2+ channel agonist, BAY K 8644 (1 μM). BAY K 8644 had an effect opposite to that of nifedipine on SEOs. The amount of depolarization during SEOs increased from 0.13 ± 0.03 to 0.45 ± 0.06 mV/s (P < 0.05), and rapidly rising oscillations appeared at the peaks of SEOs in four of five muscles (Fig.5). The large contraction induced by this activity made it difficult to maintain impalements. BAY K 8644 did not significantly change resting membrane potential (−68 ± 2.2 to −69.3 ± 1.7 mV) but reduced the frequency of SEOs from 1.8 ± 0.2 to 0.9 ± 0.1 cycles/min (P < 0.05,n = 5).
Effect of low Ca2+ solution on SEOs.
Colonic slow waves are reduced by lowering extracellular Ca2+ from 2.5 to 0.1 mM (19). Therefore, we tested the effects of reduced extracellular Ca2+ on SEOs. Reduction of bath Ca2+ from 2.5 mM to 0.1 mM completely abolished SEOs (Fig. 6 A). As SEOs were inhibited, there was a significant decrease in the membrane potential between SEOs from −66 ± 1 to −53.5 ± 2.4 mV (P < 0.5, n = 10).
Effect of Ni2+ on SEOs.
We also tested the effects of Ni2+ on SEOs, because this cation has previously been shown to inhibit colonic slow waves (19). Ni2+ (0.5 mM) completely blocked SEOs (Fig. 6 B). This effect was accompanied by an 18-mV depolarization of the membrane potential between SEOs (from −67.5 ± 2.2 to −49.6 ± 1.9 mV, n = 9,P < 0.5), but as shown above, this degree of depolarization does not block SEOs (see Fig. 2).
Effect of CPA on SEOs.
Recent studies have suggested that release of Ca2+ from intracellular stores in ICCs is responsible for the initiation of slow wave activity (16, 18), and others have shown that CPA (10 μM) reduces the frequency of slow waves in the canine colon (7). We tested the effects of CPA (10 μM) and found that this compound completely blocked SEOs (Fig.7). The inhibition was accompanied by a small depolarization in the membrane potential (e.g., from −64 ± 1.6 to −59 ± 2.1 mV, n = 7, P < 0.05).
Effect of mitochondrial inhibitors on SEOs.
Mitochondrial uncouplers have been shown to block electrical slow waves in pacemaker ICCs, and we have previously suggested that mitrochondrial uptake of Ca2+ is an essential step in generation of pacemaker currents in these cells (18). Additional experiments were performed to test whether SEOs exhibit a similar sensitivity to mitochondrial inhibitors and uncouplers. Antimycin (an inhibitor of Complex 3 in the mitochondrial electron transport chain) and FCCP (a protonophore) both reduce the electrochemical gradient necessary for Ca2+ uptake by mitochondria. Addition of antimycin (10 μM) abolished SEOs (Fig.8 A) but had no significant effect on membrane potential (−67 ± 2.7 control vs. −67 ± 2.5 mV antimycin, n = 6). FCCP (1 μM) entirely abolished SEOs (Fig. 7 B) and produced a prominent depolarization of membrane potential (from −66 ± 1.9 to −39 ± 3.7 mV, n = 4, P < 0.05; Fig. 8 B). It is unlikely that the effects of antimycin and FCCP were due to reduction in the capacity of the tissues to generate ATP, because the mitochondrial ATPase inhibitor oligomycin (10 μM, n = 5) did not affect either SEO area (0.3 ± 0.1 control vs. 0.3 ± 0.1 mV/s oligomycin), frequency (1.1 ± 0.04 control vs. 1.1 ± 0.1 cycles/min oligomycin), or resting membrane potential (−65 ± 2.6 control vs. −67 ± 2.5 mV oligomycin) (Fig. 8 C).
There is a pacemaker at the edge of the circular muscle layer that generates 16-cycles/min electrical oscillations referred to as MPOs. MPOs persist in the presence of TTX, atropine, and l-NNA (3, 11), suggesting that they are nonneural in origin. We previously reported that nitrergic stimulation radically changes the MPO pattern of activity, and the activity that develops is reminiscent of slow waves recorded near the pacemaker region at the submucosal surface of the circular muscle layer (3). We have termed the activity generated by nitrergic stimulation SEOs. The present study characterized some of the properties and pharmacology of SEOs, and the data show that SEOs are similar to slow waves. Thus nitrergic stimulation of a pacemaker in the myenteric region of the proximal colon generates electrical activity similar to the slow wave activity generated by the submucosal pacemaker.
Some of the properties SEOs share with slow waves include:1) SEOs and slow waves depend on discrete pacemaker regions at the myenteric and submucosal surfaces of the circular muscle layer, respectively. Surgical removal of these pacemaker regions blocks both events. 2) Both types of activities decline in amplitude as a function of distance from the pacemaker regions, suggesting that neither activity is actively regenerated within circular muscle cells.3) Both events are resistant to blockade by nifedipine.4) Both events are blocked by Ni2+ reduced and by extracellular Ca2+. 5) Both events are inhibited by the CaATPase inhibitor, cyclopiazonic acid (7). 6) Both events are blocked by mitochondrial inhibitors and uncouplers, such as antimycin and FCCP (18). 7) Neither event is affected by oligomycin. From these comparisons, we suggest that SEOs and slow waves may be due to similar mechanisms.
A major finding of this study is that nitrergic stimulation acts through a cGMP-dependent mechanism to suppress MPOs and generate SEOs. The mechanisms responsible for MPOs and SEOs are quite different. MPOs occur spontaneously and have different waveforms than do SEOs. MPOs are blocked by nifedipine, whereas SEOs are not. Dependence on L-type Ca2+ channels suggests that a voltage-dependent mechanism is responsible for MPOs. Switching on of SEOs appears to require cGMP. The mechanism by which cGMP initiates SEOs in myenteric pacemaker cells is not understood at the present time.
Significant evidence has accumulated suggesting that ICCs are responsible for generation of slow waves in gastrointestinal muscles (10). Networks of ICCs are found at both the myenteric and submucosal surfaces of the circular muscle layer. There is strong evidence that submucosal ICCs are responsible for the generation of slow waves, because these events are eliminated when this region is surgically removed or damaged by chemicals (8, 12, 17). Cells with morphology similar to submucosal ICCs are found at the myenteric surface of the circular muscle layer (1, 15), and there has been speculation that myenteric ICCs could provide pacemaker activity responsible for MPOs. It is possible that SEOs also could be generated by ICCs located within the myenteric plexus region. This hypothesis has not yet been directly tested. However, surgical removal of both submucosal and myenteric borders of the circular muscle layer renders the remaining muscle quiescent (9).
Recent studies suggest that rhythmic inward (pacemaker) currents in cultured ICCs are due to cation channels permeable to Ca2+and Na+ (5). These channels are voltage independent and insensitive to block by L-type Ca2+ channel inhibitors, such as nifedipine. The pacemaker currents are blocked by divalent cations, such as gadolinium (5). The sensitivity of pacemaker current to divalent cations and insensitivity to nifedipine is similar to the effects of these agents on slow waves in intact muscles [slow waves are resistant to nifedipine but are blocked by Ni2+ (19)]. In the present study, SEOs were resistant to nifedipine and abolished by Ni2+. Thus it is possible that SEOs may be generated by nonselective cation currents similar to the currents observed in ICCs.
Although Ni2+ blocks both SEOs and slow waves, there were significant differences in the effect of Ni2+ on resting membrane potential. Ni2+ blockade of SEOs is accompanied by a large depolarization of resting membrane potential. Previous experiments in which slow waves were recorded from cells near the submucosal surface of the circular muscle layer did not find that Ni2+ caused depolarization (19). Mechanisms responsible for the depolarization of cells near the myenteric surface (this study) were not determined, but the difference noted suggests that different conductances may contribute to resting potentials in the two regions of the circular muscle layer. Depolarization per se was unlikely to cause the block of SEOs by Ni2+, because an equivalent depolarization with elevated extracellular K+did not abolish SEOs (Fig. 2 A).
Nifedipine did not abolish SEOs, but it significantly increased the frequency of these events. Nifedipine had similar effects on slow waves (19). In contrast, the L-type Ca2+ channel agonist BAY K 8644 decreased the frequency of both SEOs and slow waves (19). Changes in slow wave and SEO frequency suggest that Ca2+ influx via L-type Ca2+ channels modulate the processes that control pacemaker activity, and previous studies have reported that L-type Ca2+ currents are expressed by ICCs from the canine colon (6).
Reducing extracellular Ca2+ from 2.5 to 0.1 mM or the addition of Ni2+ blocked myenteric SEOs (this study) and slow waves (19). Although Na+ is likely to be the primary charge carrier through the pacemaker conductance (5), blockade of activity by procedures that would tend to inhibit Ca2+ influx suggests that Ca2+ entry is necessary for maintaining rhythmicity. Ca2+ entry may be important for maintaining Ca2+ stores that also appear to be critical for slow waves (7, 13, 18) and SEO (this study) generation.
Recent studies have linked pacemaker current generation in ICCs to coupling between inositol 1,4,5-trisphosphate-dependent Ca2+ release from the endoplasmic reticulum and Ca2+ uptake into adjacent mitochondria (18). Tests in the present study using mitochondrial inhibitors and uncouplers suggest that a similar mechanism may drive SEOs. Antimycin abolished SEOs without producing a significant change in resting membrane potential. Block of an SEO was unlikely to be due to ATP depletion, because oligomycin, an inhibitor of the F0/F1 ATPase, did not affect SEOs. SEOs were also blocked by FCCP, and this effect was accompanied by depolarization.
The most prominent feature that distinguished SEOs from slow waves was the dependence of SEOs on nitrergic stimulation (3). Because the actions of SNP were abolished by the guanylyl cyclase inhibitor ODQ, it is likely that SNP acts via generation of cGMP. In contrast, slow waves occur spontaneously in cells near the submucosal pacemaker, and concentrations of SNP that greatly modified the pattern of activity in the myenteric half of the circular muscle produced only a modest inhibition of slow waves in cells near the submucosal surface (3). Interestingly, cGMP has been shown to inhibit pacemaker currents in cultured ICCs from the murine small intestine (4). This action is more in line with the effects of NO on slow waves than the stimulation of SEOs by SNP/cGMP. At present, the cGMP-dependent regulation of pacemaker activity in the myenteric region is not understood, but effects of the cGMP pathway on Ca2+handling mechanisms may be involved.
In summary, our results suggest that there are pacemakers at the myenteric surface of the circular muscle layer capable of generating two discrete forms of electrical rhythmicity, MPOs and SEOs. MPOs (nifedipine-sensitive) occur spontaneously in the majority of preparations, and SEOs (nifedipine-resistant) are switched on via nitrergic stimulation via a cGMP-dependent mechanism. SEOs are a specialized type of electrical slow wave with properties very similar to the electrical slow waves generated at the submucosal surface of the circular muscle layer. Because the myenteric pacemaker region also contains networks of ICCs, it is possible that SEOs are generated by myenteric ICCs after activation by NO/cGMP.
This project was funded by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-41315.
Address for reprint requests and other correspondence: K. D. Keef, Dept. of Physiology and Cell Biology, Univ. of Nevada School of Medicine, Reno, NV 89557 ().
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.
- Copyright © 2002 the American Physiological Society