Characterization of slow waves generated by myenteric interstitial cells of Cajal of the rabbit small intestine

Abstract

Slow waves (slow waves_ICC) were recorded from myenteric interstitial cells of Cajal (ICC-MY) in situ in the rabbit small intestine, and their properties were compared with those of mouse small intestine. Rabbit slow waves_ICC consisted of an upstroke depolarization followed by a distinct plateau component. Ni²⁺ and nominally Ca²⁺-free solutions reduced the rate-of-rise and amplitude of the upstroke depolarization. Replacement of Ca²⁺ with Sr²⁺ enhanced the upstroke component but decreased the plateau component of rabbit slow waves_ICC. In contrast, replacing Ca²⁺ with Sr²⁺ decreased both components of mouse slow waves_ICC. The plateau component of rabbit slow waves_ICC was inhibited in low-extracellular-Cl⁻-concentration (low-[Cl⁻]_o) solutions and by 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS), an inhibitor of Cl⁻ channels, cyclopiazonic acid (CPA), an inhibitor of internal Ca²⁺ pumps, or bumetanide, an inhibitor of Na⁺-K⁺-2Cl⁻ cotransporter (NKCC1). Bumetanide also inhibited the plateau component of mouse slow waves_ICC. NKCC1-like immunoreactivity was observed mainly in ICC-MY in the rabbit small intestine. Membrane depolarization with a high-K⁺ solution reduced the upstroke component of rabbit slow waves_ICC. In cells depolarized with elevated external K⁺, DIDS, CPA, and bumetanide blocked slow waves_ICC. These results suggest that the upstroke component of rabbit slow waves_ICC is partially mediated by voltage-dependent Ca²⁺ influx, whereas the plateau component is dependent on Ca²⁺-activated Cl⁻ efflux. NKCC1 is likely to be responsible for Cl⁻ accumulation in ICC-MY. The results also suggest that the mechanism of the upstroke component differs in rabbit and mouse slow waves_ICC in the small intestine.

interstitial cells of cajal (ICC) are distributed throughout the gastrointestinal (GI) tract in mammalian species, including humans (30, 31, 35, 40, 41, 43, 44). ICC express Kit, a receptor tyrosine kinase, encoded at the W locus (32, 35). The ligand for Kit is stem cell factor (SCF), encoded at the steel locus (Sl). Development, differentiation, and maintenance of ICC depend on SCF-Kit signaling pathway (36, 47). After the development of techniques to label ICC (35, 45) with monoclonal antibodies for Kit (ACK2) (31), most studies to evaluate the function of ICC focused on the mouse small intestine.

In Kit mutants (W/W^v mice) (14, 45) or SCF mutants (Sl/Sl^d mice) (46), myenteric ICC (ICC-MY) were largely missing from the small intestine. Slow wave activity was lost in the small intestines of these mutants (14, 45, 46). Therefore, it is likely that slow waves (pacemaker activity) originate in ICC-MY in the small intestine (37). Direct recording of electrical activity from ICC-MY in the mouse small intestine in situ showed that ICC-MY generate large rhythmic potential changes (slow waves_ICC), of which amplitude and maximum rate-of-rise (dV/dt_max) are greater than when slow waves are recorded from circular smooth muscle cells (CSMCs) (20–23). Mouse slow waves_ICC consist of at least two major components: a rapid upstroke component followed by a plateau component (20, 21). The upstroke of mouse slow waves_ICC is initiated by nifedipine-resistant, voltage-dependent Ca²⁺ influx, whereas the plateau is due to Ca²⁺-activated Cl⁻ efflux (20). A recent study has demonstrated that anoctamin 1 (Ano1) protein, which functions as a Ca²⁺-activated Cl⁻ conductance (CaCC), is expressed with high specificity in ICC of mouse, monkey, and human GI tracts (6, 15, 51).

In rabbit, cat, dog, monkey, and human intestinal muscles, CSMCs exhibit slow waves at frequencies between 8–15 min⁻¹, which is slower than the frequency of slow waves recorded from mouse CSMCs (i.e., 30–40 min⁻¹) (2, 3, 8, 9, 15, 20, 23). Hara et al. (8) suggested that “slow waves of the small intestine of the rabbit, cat, dog and human, are generated in nonneural cells located between the longitudinal and outer circular muscle layers.” The nonneural cells may be myenteric ICC (ICC-MY) because these cells reside in the vicinity where pacemaker activity originates and isolated ICC-MY of the intestine display spontaneous pacemaker activity (31, 35, 36, 51). There are no direct data, however, demonstrating that ICC-MY generate slow waves in the small intestines of larger animals and humans.

The present study describes the properties of slow waves recorded from ICC-MY of the rabbit small intestine in situ. Slow waves recorded from ICC-MY (slow waves_ICC) consisted of two components: a rapidly rising, upstroke component and a plateau component. A hypothesis is presented suggesting that Cl⁻ handling mechanisms are involved in the generation of the plateau component of slow waves_ICC. Similarities and differences in the properties of slow waves_ICC recorded from rabbit and mouse intestines are discussed. Some of the results were reported briefly in the VII^th International Symposium on Interstitial Cells of Cajal (25).

MATERIALS AND METHODS

Tissue preparations.

Male Japanese albino rabbits, weighing 2–2.5 kg, were anesthetized by injection of pentobarbitone sodium (50 mg/kg, given iv) and killed by exsanguination. BALB/c mice of either sex, aged 4–6 wk, were anesthetized with fluoromethyl 2,2,2-trifluoro-1-(trifluoromethyl) ethyl ether (sevoflurane, Maruishi Pharmaceutical, Osaka, Japan) and killed by cervical dislocation and exsanguination. Maintenance of animals and all experimental protocols involving animals were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of Nagoya City University Medical School and accredited by The Physiological Society of Japan. All procedures were approved by Nagoya City University Animal Center (approval no. H18M-31). Segments of terminal ileum were removed and opened along the mesenteric border. The mucosal and serosal layers were carefully removed while the tissues were viewed with a dissecting microscope. Small segments of muscle (∼1 × 1 mm for rabbit; ∼0.5 × 0.5 mm for mouse) were pinned with tungsten wires (φ0.03 mm) on a silicone rubber plate with the longitudinal muscle uppermost. The plate was fixed to the bottom of an organ bath (8 mm wide, 8 mm deep, 20 mm long), and the bath was continuously perfused (2 ml/min) with warmed (35°C) and oxygenated Krebs solution. Experiments were performed in the presence of nifedipine (3 μM) to minimize muscle movements. Nifedipine (3 μM) had no effect on the electrophysiological parameters of slow waves_ICC in the rabbit small intestine (see results).

Electrophysiological recording.

Conventional microelectrode techniques were used to record intracellular electrical activity from smooth muscle tissues. Microelectrodes were pulled from glass capillaries (outer diameter 1.5 mm, inner diameter 0.86 mm) and filled with 2 M KCl. The electrodes had tip resistances between 50 and 80 MΩ. Electrical responses were recorded via a high-input impedance amplifier (Axoclamp-2B, Axon Instruments, Foster City, CA), displayed on a cathode-ray oscilloscope (SS-7602, Iwatsu, Osaka, Japan), and stored on a PC-style computer for later analysis and display. Electrophysiological parameters of slow waves_ICC recorded during the impalements of cells were evaluated by tabulating the properties of individual slow waves over at least 10 slow wave cycles. Values of slow waves_ICC parameters reported in the text were determined by calculating means and standard deviations for impalements of several ICC in different animals (n > 4).

The morphological features of cells impaled in ileal muscles were identified by filling the cells with 0.5% (wt/vol) propidium iodide added to pipet solution (PI; Sigma, St. Louis, MO). Impaled cells were filled with PI by passing hyperpolarizing current pulses (duration 100 ms, intensity 1 nA, frequency 3 Hz for 5–30 min) supplied by an electric stimulator (SEN-3301, Nihon Kohden, Tokyo, Japan) (24, 26). After filling, the muscles were fixed overnight at 4°C with fresh 4% (wt/vol) paraformaldehyde in 0.1 M phosphate-buffered saline (PBS). After fixation, the muscles were washed several times with PBS, mounted in Dako fluorescent mounting medium (Dako), covered with a coverslip, and viewed with a confocal microscope (LSM5 PASCAL, Carl Zeiss). A confocal microscope with a krypton-argon laser was used to visualize cells filled with propidium iodide (488 nm excitation filter and 560 nm emission long-pass filter).

Immunohistochemical studies.

Segments of rabbit terminal ileum were removed and immersed in PBS maintained at 4°C. The tissue was cut along the mesenteric border, and the mucosa and a part of the circular muscle layer were removed with sharp tweezers to obtain whole mount preparations of the longitudinal muscle layer. The preparations were flattened, pinned, and immersed in acetone for 15 min at room temperature. The fixed whole mount preparations were washed twice in PBS (5 min each).

All primary antibodies used in this study were diluted in PBS containing 2% bovine serum albumin (BSA), 0.3% Triton X-100, and 0.01% sodium azide. All secondary antibodies were diluted in PBS containing 2% BSA. Antibodies used were as follows: goat polyclonal antibody for Na⁺-K⁺-2Cl⁻ cotransporter (NKCC1; 1:50, Santa Cruz Biotechnology), mouse monoclonal anti-vimentin antibody (1:50, clone V9, Dako), goat polyclonal anti-Kit antibody (1:50, M-14, Santa Cruz Biotechnology), tetramethylrhodamine isothiocyanate (TRITC)-conjugated donkey anti-goat Ig antibody (1:100, Chemicon) and fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse Ig antibody (1:100, Dako).

The whole mounts were incubated with 0.3% Triton X-100 in PBS for 10 min, incubated with Block Ace (Dainippon Seiyaku) for 20 min at room temperature and incubated with primary antibodies for 2 days at 4°C. Whole mounts were washed in PBS and incubated with secondary antibodies for 2 h at room temperature. No immunoreactivity was detected in preparations for which primary antibodies were not used. The specimens were examined under a confocal laser scanning microscope (LSM 5 PASCAL, Carl Zeiss).

Statistics.

Experimental values were expressed by the mean value ± SD. Statistical significance was tested by Student's t-test, and probabilities of less than 5% (P < 0.05) were considered significant.

Solutions and drugs.

The ionic composition of the Krebs solution was as follows (in mM): 137.4 Na⁺, 5.9 K⁺, 2.5 Ca²⁺, 1.2 Mg²⁺, 15.5 HCO₃⁻, 1.2 H₂PO₄⁻, 134 Cl⁻, and 11.5 glucose. Solution containing high-potassium ion concentration [high-K⁺ solution; extracellular K⁺ concentration ([K⁺]_o) = 20.0 mM] was prepared by replacing NaCl with KCl. Low-Cl⁻ solution [extracellular Cl⁻ concentration ([Cl⁻]_o) = 13.3 mM] was prepared by equimolar replacement of NaCl with sodium isethionate. The solutions were aerated with O₂ containing 5% CO₂, and the pH of the solutions was maintained at 7.2–7.3.

Drugs used were bumetanide, cyclopiazonic acid (CPA), 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS), furosemide, nifedipine, niflumic acid (NFA), tetrodotoxin (TTX) (all from Sigma). CPA, bumetanide, furosemide, and nifedipine were dissolved in dimethyl sulfoxide (DMSO) to make stock solutions and were added to Krebs solution to make the desired concentrations, just prior to use. All other drugs were dissolved first in distilled water. The final concentration of the solvent in Krebs solution did not exceed 1/1,000. Addition of these chemicals to Krebs solution did not alter the pH of the solution.

RESULTS

General observations.

When intracellular recordings were made from ICC-MY of the rabbit small intestine, rapidly rising, rhythmical membrane potential changes (slow waves_ICC) were detected (Fig. 1A, Table 1). In 10 preparations, recording of slow waves from impalements of ICC-MY was confirmed by filling cells with fluorescent dye during recording (26). Figure 1B shows a confocal image of a rabbit slow waves_ICC-generating cell visualized by filling with propidium iodide. The slow waves_ICC-generating cells had fusiform cell bodies and multiple processes, showing the typical morphology of ICC-MY (30, 31, 41). Rabbit slow waves_ICC had two components: a rapidly rising upstroke component followed by a plateau component. The amplitude of the upstroke component was more than 50 mV (Table 1). The maximum rate-of-rise (dV/dt_max) of the upstroke component was ∼10 Vs⁻¹ (Table 1). Rabbit slow waves_ICC occurred at a frequency of ∼10 min⁻¹ and had a half-width (measured at 50% peak amplitude of the plateau component) of 2–3 s (Table 1). The resting membrane potential (RMP) of rabbit small intestinal ICC-MY was −63 to −71 mV (Table 1). In the mouse small intestine, ICC-MY generated slow waves_ICC with amplitudes in excess of 50 mV and dV/dt_max less than 3 Vs⁻¹ (Fig. 1C, Table 1). Mouse slow waves_ICC occurred at a frequency of ∼25 min⁻¹ and had a half-width of less than 1 s (Fig. 1C, Table 1). The RMP of mouse small intestinal ICC-MY was −64 to −73 mV (Table 1). Rabbit slow waves_ICC were also characterized by the presence of a rapid repolarization just after the upstroke component (Fig. 1, A and D), which was not observed in the waveforms of mouse slow waves_ICC (Fig. 1, C and D). Compared with mouse slow waves_ICC, rabbit slow waves_ICC had larger dV/dt_max, slower frequency, and longer half-width. Smooth muscle layers in the rabbit small intestine are thicker than those in mouse small intestine. Hence, the rapidly rising, upstroke component of rabbit slow waves_ICC may represent a suitable mechanism to increase synchrony of membrane depolarization through the smooth muscle layers. These results suggest that the cellular mechanisms underlying the generation of the upstroke component of rabbit slow waves_ICC might be different from those of mouse slow waves_ICC.

Fig. 1.

Rabbit slow waves recorded from interstitial cells of Cajal (slow waves_ICC) and mouse slow waves_ICC. A: slow waves recorded from myenteric interstitial cells of Cajal (ICC-MY) in the rabbit small intestine. Amplitude of upstroke component (Aa), amplitude of plateau component (Ab), half-width (Ac), and interval between slow waves_ICC (Ad) are shown. B: confocal image of a rabbit slow waves_ICC-generating cell injected with propidium iodide shown in A. The injected part is marked by an arrowhead. Scale bar in B represents 50 μm. C: slow waves recorded from ICC-MY in the mouse small intestine. Amplitude of upstroke component (Ca), half-width (Cb), and interval between slow waves_ICC (Cc) are shown. D: comparison of expanded trace of rabbit slow waves_ICC (Da) with that of mouse slow waves_ICC (Db). The resting membrane potentials (dotted lines) were A, −67 mV; C, −65 mV.

Table 1.

Comparison of properties of slow waves_ICC in the rabbit and mouse small intestine

	Membrane potential, mV	Amplitude (upstroke), mV	Amplitude (plateau), mV	Frequency, min−1	Half-width, s	dV/dtmax, V/s	n
Rabbit	−66 ± 4	56 ± 5	50 ± 4	10.6 ± 1.8	2.47 ± 0.32	10.7 ± 1.7	38
Mouse	−67 ± 3	55 ± 5		24.4 ± 1.9†	0.81 ± 0.08†	2.2 ± 0.3†	22

Values are means ± SD; n = number of animals. slow waves_ICC, Slow waves recorded from interstitial cells of Cajal; dV/dt_max, maximum rate-of-rise.

^*P < 0.05,

^†P < 0.01, compared with rabbit.

Effects of NiCl₂ and nominally Ca²⁺-free solution on slow waves_ICC.

Application of NiCl₂, Ca²⁺-free solution, or mibefradil, a blocker of voltage-dependent T-type Ca²⁺ channel, reduced the magnitude and rate-of-rise (dV/dt_max) of the upstroke component of slow waves_ICC recorded from ICC-MY in the mouse small intestine, indicating that voltage-dependent Ca²⁺ influx may generate the upstroke component (20, 21). To characterize the upstroke component of rabbit slow waves_ICC, the effects of NiCl₂ and nominally Ca²⁺-free solution on slow waves_ICC were tested. NiCl₂ (60 μM) decreased the amplitude of the upstroke component (control, 59 ± 4 mV; in NiCl₂, 53 ± 2 mV; n = 5; P < 0.05), frequency (control, 10.6 ± 1.6 min⁻¹; in NiCl₂, 7.9 ± 1.8 min⁻¹; n = 5; P < 0.05) and dV/dt_max (control, 11.0 ± 1.9 Vs⁻¹; in NiCl₂, 6.2 ± 1.5 Vs⁻¹; n = 5; P < 0.05) of slow waves_ICC, with no alteration to the amplitude of the plateau component (control, 52 ± 4 mV; in NiCl₂, 53 ± 3 mV; n = 5; P > 0.05) or half-width (control, 2.5 ± 0.3 s; in NiCl₂, 2.6 ± 0.5 s; n = 5; P > 0.05) (Fig. 2, A and B). NiCl₂ (60 μM) had no effect on RMP. Nifedipine (3 μM) had no effect on the amplitude of the upstroke component (control, 57 ± 1 mV; in nifedipine, 57 ± 2 mV; n = 4; P > 0.05), amplitude of the plateau component (control, 52 ± 1 mV; in nifedipine, 52 ± 2 mV; n = 4; P > 0.05), frequency (control, 12.3 ± 1.7 min⁻¹; in nifedipine, 12.0 ± 1.3 min⁻¹; n = 4; P > 0.05), half-width (control, 2.1 ± 0.1 s; in nifedipine, 2.1 ± 0.3 s; n = 4; P > 0.05), or dV/dt_max (control, 10.4 ± 0.6 Vs⁻¹; in nifedipine, 10.8 ± 2.5 Vs⁻¹; n = 4; P > 0.05) of slow waves_ICC, suggesting that voltage-dependent L-type Ca²⁺ channel is not involved in the generation of slow waves_ICC in the rabbit small intestine. Nifedipine (3 μM) also had no effect on RMP. Application of nominally Ca²⁺-free solution hyperpolarized the membrane by 4 ± 1 mV transiently (2.7 ± 1.1 min) and depolarized cells by 4 ± 3 mV (n = 6) after 11.0 ± 2.2 min (Fig. 2C). Nominally Ca²⁺-free solution decreased the amplitude of the upstroke component (control, 56 ± 5 mV; in nominally Ca²⁺-free solution, 40 ± 5 mV; n = 6; P < 0.01), amplitude of the plateau component (control, 51 ± 3 mV; in nominally Ca²⁺-free solution, 38 ± 6 mV; n = 6; P < 0.01), frequency (control, 11.6 ± 1.7 min⁻¹; in nominally Ca²⁺-free solution, 3.3 ± 1.8 min⁻¹; n = 6; P < 0.01), half-width (control, 2.5 ± 0.2 s; in nominally Ca²⁺-free solution, 1.9 ± 0.3 s; n = 6; P < 0.01) and dV/dt_max (control, 10.2 ± 1.3 Vs⁻¹; in nominally Ca²⁺-free solution, 3.0 ± 1.4 Vs⁻¹; n = 6; P < 0.01) of slow waves_ICC (Fig. 2, C and D). These results suggest that the upstroke component of rabbit slow waves_ICC is partially dependent on Ca²⁺ influx. The results also show that slow waves_ICC of the rabbit small intestine persist in very low extracellular Ca²⁺.

Fig. 2.

Effects of NiCl₂ and nominally Ca²⁺-free solution on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during application of 60 μM NiCl₂ (Ab). B: expanded traces of rabbit slow waves_ICC recorded in the absence (Ba) and presence of 60 μM NiCl₂ (Bb). C: rabbit slow waves_ICC were recorded before (Ca) and during application of nominally Ca²⁺-free solution (Cb). D: expanded traces of rabbit slow waves_ICC recorded in the absence (Da) and presence of nominally Ca²⁺-free solution (Db). The resting membrane potentials (dotted lines) were A, −68 mV; C, −65 mV. A and C were recorded from different tissues.

Effects of Ca²⁺ replacement with Sr²⁺ on slow waves_ICC.

Since Sr²⁺ can replace Ca²⁺ and produce Sr²⁺ influx through various types of Ca²⁺ channels, Sr²⁺ has been used as the charge carrier to study the properties of voltage-dependent Ca²⁺ channels (11, 33). To test the possibility that Sr²⁺ might replace Ca²⁺ in the generation of the upstroke component of rabbit slow waves_ICC, extracellular Ca²⁺ (2.5 mM) was replaced with Sr²⁺ (2.5 mM). Replacement of Ca²⁺ with Sr²⁺, increased the amplitude of the upstroke component and dV/dt_max, but the amplitude of the plateau component, frequency, and half-width decreased (Fig. 3, A and B, Table 2), suggesting that Ca²⁺ replacement with Sr²⁺ enhanced the upstroke component but inhibited the plateau component of slow waves_ICC. In contrast, Ca²⁺ replacement with Sr²⁺ inhibited both components of mouse slow waves_ICC (Fig. 3C and D, Table 2).

Fig. 3.

Effects of Sr²⁺ on rabbit and mouse slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during replacement of Ca²⁺ with Sr²⁺ (Ab). B: expanded traces of rabbit slow waves_ICC recorded before (Ba) and after substitution of Sr²⁺ for Ca²⁺ (Bb). C: mouse slow waves_ICC were recorded before (Ca) and during replacement of Ca²⁺ with Sr²⁺ (Cb). D: expanded traces of mouse slow waves_ICC recorded before (Da) and after substitution of Sr²⁺ for Ca²⁺ (Db). The resting membrane potentials (dotted lines) were A, −67 mV; C, −63 mV.

Table 2.

Effects of SrCl₂ on slow waves_ICC in the rabbit and mouse small intestine

	Membrane potential, mV	Amplitude (upstroke), mV	Amplitude (plateau), mV	Frequency, min−1	Half-width, s	dV/dtmax, V/s	n
Rabbit
Control	−65 ± 3	58 ± 4	52 ± 4	10.2 ± 2.5	2.63 ± 0.41	9.7 ± 0.8	7
SrCl2	−65 ± 3	62 ± 5*	49 ± 6*	4.2 ± 2.1†	1.75 ± 0.15†	15.5 ± 1.6*	7
Mouse
Control	−69 ± 5	59 ± 6		25.5 ± 2.3	0.81 ± 0.11	2.6 ± 0.3	7
SrCl2	−67 ± 5	58 ± 7		8.2 ± 3.1†	0.58 ± 0.05†	1.2 ± 0.4†	7

Values are means ± SD; n = number of animals.

^*P < 0.05,

^†P < 0.01, compared with control.

To further study the effects Sr²⁺ on the upstroke component, Sr²⁺ was applied to nominally Ca²⁺-free solution during the recording of rabbit slow waves_ICC. In the presence of nominally Ca²⁺-free solution (15 min), Sr²⁺ (2.5 mM) increased the amplitude (Ca²⁺-free solution, 40 ± 3 mV; Sr²⁺, 58 ± 5 mV; n = 4; P < 0.05) and dV/dt_max (Ca²⁺-free solution, 3.0 ± 1.2 Vs⁻¹; Sr²⁺, 15.1 ± 1.5 Vs⁻¹; n = 4; P < 0.01) of slow waves_ICC, whereas Sr²⁺ (2.5 mM) decreased the frequency (Ca²⁺-free solution, 3.7 ± 0.6 min⁻¹; Sr²⁺, 2.2 ± 0.4 min⁻¹; n = 4; P < 0.05) of slow waves_ICC (Fig. 4). These results suggest that Sr²⁺ serves as a suitable charge-carrier for the conductance responsible for the upstroke component of slow waves_ICC of the rabbit small intestine.

Fig. 4.

Effects of Sr²⁺ on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before and during application of nominally Ca²⁺-free solution (sol.) for 15 min, and 2.5 mM SrCl₂ was added as indicated by the bar under the record. The resting membrane potential (dotted line) was −64 mV. B: expanded traces of slow waves_ICC recorded before (Ba), during application of nominally Ca²⁺-free solution over 15 min (Bb), and during coapplication of 2.5 mM SrCl₂ (Bc) as indicated in A.

Effects of low-[Cl⁻]_o solution on slow waves_ICC.

Reducing [Cl⁻]_o has been reported to inhibit the plateau component of guinea pig gastric slow waves_ICC (18) and mouse small intestinal slow waves_ICC (20). Reducing [Cl⁻]_o from 134 mM to 13.3 mM increased the amplitude of the upstroke component (control, 56 ± 4 mV; low-[Cl⁻]_o, 60 ± 3 mV; n = 6; P < 0.05), amplitude of the plateau component (control, 51 ± 1 mV; low-[Cl⁻]_o, 58 ± 3 mV; n = 6; P < 0.01), and frequency (control, 9.5 ± 1.1 min⁻¹; low-[Cl⁻]_o, 11.6 ± 1.8 min⁻¹; n = 6; P < 0.05) of slow waves_ICC transiently (0.8 ± 0.2 min), with a decrease of the half-width (control, 2.6 ± 0.4 s; low-[Cl⁻]_o, 2.0 ± 0.3 s; n = 6; P < 0.05) and no alteration in dV/dt_max (control, 10.3 ± 1.4 Vs⁻¹; low-[Cl⁻]_o, 11.6 ± 2.8 Vs⁻¹; n = 6; P > 0.05) (Fig. 5, A, Ba, and Bb). After 6–9 min (7.2 ± 0.7 min) in reduced [Cl⁻]_o, the amplitude of the plateau component (control, 51 ± 1 mV; low-[Cl⁻]_o, 47 ± 1 mV; n = 6; P < 0.01), frequency (control, 9.5 ± 1.1 min⁻¹; low-[Cl⁻]_o, 7.8 ± 1.1 min⁻¹; n = 6; P < 0.05), and half-width (control, 2.6 ± 0.4 s; low-[Cl⁻]_o, 1.2 ± 0.1 s; n = 6; P < 0.01) of slow waves_ICC decreased. Reducing [Cl⁻]_o caused no significant change in the amplitude of the upstroke component (control, 56 ± 4 mV; low-[Cl⁻]_o, 53 ± 3 mV; n = 6; P > 0.05) or dV/dt_max (control, 10.3 ± 1.4 Vs⁻¹; low-[Cl⁻]_o, 9.5 ± 1.6 Vs⁻¹; n = 6; P > 0.05) (Fig. 5, A, Ba, Bc). These results suggest that reducing [Cl⁻]_o causes a biphasic response in the plateau component of rabbit intestinal slow waves_ICC.

Fig. 5.

Effects of low-[Cl⁻]_o solution on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before and during application of low-[Cl⁻]_o solution. Low-[Cl⁻]_o solution was added as indicated by the bar under the record. The resting membrane potential (dotted lines) was −69 mV. B: expanded traces of rabbit slow waves_ICC recorded before (Ba) and during application of low-[Cl⁻]_o solution for 0.5 min (Bb) and 7 min (Bc) as indicated in A.

Effects of Cl⁻ channel blockers on slow waves_ICC.

DIDS, a Cl⁻ channel blocker, abolished the plateau component of guinea pig gastric slow waves_ICC (18, 19) and mouse small intestinal slow waves_ICC (20). We tested the effects of DIDS on rabbit slow waves_ICC. As shown in Fig. 6A, DIDS (500 μM) decreased the amplitude of the plateau component (control, 48 ± 2 mV; in DIDS, 35 ± 3 mV; n = 6; P < 0.01), half-width (control, 2.4 ± 0.3 s; in DIDS, 1.0 ± 0.1 s; n = 6; P < 0.01) and dV/dt_max (control, 9.6 ± 1.0 Vs⁻¹; in DIDS, 6.1 ± 0.1 Vs⁻¹; n = 6; P < 0.05) of slow waves_ICC and increased the frequency (control, 12.8 ± 0.7 min; in DIDS, 18.2 ± 0.7 min; n = 6; P < 0.01), without affecting the amplitude of the upstroke component (control, 52 ± 1 mV; in DIDS, 52 ± 2 mV; n = 6; P > 0.05) (see also Fig. 6B). DIDS (500 μM) had no effect on RMP. These results suggest that DIDS inhibited the plateau component of rabbit slow waves_ICC.

Fig. 6.

Effects of DIDS on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during application of 500 μM DIDS (Ab). B: expanded traces of rabbit slow waves_ICC were recorded in the absence (Ba) and presence of 500 μM DIDS (Bb). C: rabbit slow waves_ICC were recorded during application of 20.0 mM extracellular K⁺ concentration ([K⁺]_o) solution followed by 500 μM DIDS (shown by the bar). The resting membrane potentials (dotted lines) were A, −66 mV; C, −69 mV. A and C were recorded from different tissues.

High K⁺-solution (20.0 mM [K⁺]_o) depolarized cells (20 ± 2 mV, n = 8), abolished the upstroke component and inhibited the amplitude of the plateau component (control, 49 ± 3 mV; in 20.0 mM [K⁺]_o solution, 30 ± 2 mV; n = 8; P < 0.01), frequency (control, 9.8 ± 0.7 min⁻¹; in 20.0 mM [K⁺]_o solution, 6.9 ± 0.6 min⁻¹; n = 8; P < 0.01) and dV/dt_max (control, 9.6 ± 1.1 Vs⁻¹; in 20.0 mM [K⁺]_o solution, 0.5 ± 0.1 Vs⁻¹; n = 8; P < 0.01) of rabbit slow waves_ICC, with an increase of the half-width (control, 2.7 ± 0.5 s; in 20.0 mM [K⁺]_o solution, 5.3 ± 0.3 s; n = 8; P < 0.01) in the presence of TTX (3 μM), suggesting that 20.0 mM [K⁺]_o solution mainly affected the generation of the upstroke component of rabbit slow waves_ICC as reported previously in guinea pig gastric slow waves_ICC (18) and mouse small intestinal slow waves_ICC (20). In the presence of 20.0 mM [K⁺]_o, DIDS (500 μM) reduced slow waves_ICC to 3 ± 1 mV (n = 5) (Fig. 6C).

Effects of CPA on slow waves_ICC.

CPA, an inhibitor of internal Ca²⁺ pump, inhibited the plateau component of guinea pig gastric slow waves_ICC (18) and mouse small intestinal slow waves_ICC (22). To evaluate the role of Ca²⁺ released from internal Ca²⁺ stores in the generation of the plateau component of rabbit slow waves_ICC, the effects of CPA on rabbit slow waves_ICC were studied. CPA (10 μM) inhibited the amplitude of the upstroke component (control, 53 ± 4 mV; in CPA, 46 ± 4 mV; n = 7; P < 0.05), amplitude of the plateau component (control, 48 ± 4 mV; in CPA, 34 ± 4 mV; n = 7; P < 0.01), half-width (control, 2.6 ± 0.2 s; in CPA, 1.1 ± 0.1 s; n = 7; P < 0.01) and dV/dt_max (control, 9.6 ± 1.0 Vs⁻¹; in CPA, 7.6 ± 1.3 Vs⁻¹; n = 7; P < 0.05) of rabbit slow waves_ICC, with an increase of the frequency (control, 9.8 ± 0.7 min⁻¹; in CPA, 13.8 ± 2.4 min⁻¹; n = 7; P < 0.05) in the presence of TTX (3 μM) (Fig. 7, A and B). CPA (10 μM) depolarized the membrane by 5 ± 2 mV. High K⁺-solution (10.6 mM [K⁺]_o) depolarized the membrane (7 ± 2 mV, n = 5) in the presence of TTX (3 μM). However, this concentration of high K⁺-solution did not affect the half-width of rabbit slow waves_ICC (control, 2.3 ± 0.3 s; in 10.6 mM [K⁺]_o solution, 2.6 ± 0.2 s; n = 5; P > 0.05). Therefore, the CPA-induced inhibitory effect on the half-width of rabbit slow waves_ICC was not due to depolarization of the membrane alone. In 20.0 mM high-K⁺ solution, CPA (10 μM) reduced slow waves_ICC to 4 ± 2 mV (n = 4) when cells were depolarized (Fig. 7C), confirming that CPA has inhibitory effects on the plateau component of slow waves_ICC. These results suggest that functional internal Ca²⁺ stores are essential for the generation of the plateau component of rabbit slow waves_ICC.

Fig. 7.

Effects of cyclopiazonic acid (CPA) on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during application of 10 μM CPA (Ab). B: expanded traces of rabbit slow waves_ICC recorded in the absence (Ba) and presence of 10 μM CPA (Bb). C: rabbit slow waves_ICC were recorded during application of 20.0 mM [K⁺]_o solution followed by 10 μM CPA (shown by the bar). The resting membrane potentials (dotted lines) were A, −63 mV; C, −71 mV. A and C were recorded from different tissues.

These observations, when combined with those presented in Fig. 5 and Fig. 6, indicate that the plateau component of slow waves_ICC is generated by the activation of Ca²⁺-activated Cl⁻ channels (CaCCs) in the rabbit small intestine.

Effects of NKCC1 inhibitors on slow waves_ICC.

The Na⁺-K⁺-2Cl⁻ cotransporter (NKCC1) is strongly expressed in ICC-MY and may be involved in slow wave generation in the mouse small intestine (48). We tested the effects of a blocker of NKCC1 on rabbit slow waves_ICC. Bumetanide (10 μM), known to inhibit NKCC1 selectively (34), inhibited the amplitude of the plateau component and half-width of rabbit slow waves_ICC, and increased of the frequency. Bumetanide had no effect on the amplitude of the upstroke component or dV/dt_max (Fig. 8, A and B, Table 3). After depolarization of cells with 20.0 mM K⁺ solution, bumetanide abolished slow waves_ICC (Fig. 8C) as in the case of DIDS (Fig. 6C) or CPA (Fig. 7C).

Fig. 8.

Effects of bumetanide on rabbit slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during application of 10 μM bumetanide (Ab). B: expanded traces of rabbit slow waves_ICC recorded in the absence (Ba) and presence of 10 μM bumetanide (Bb). C: rabbit slow waves_ICC were recorded during application of 20.0 mM [K⁺]_o solution followed by 10 μM bumetanide (shown by the bar). The resting membrane potentials (dotted lines) were A, −63 mV; C, −70 mV. A and C were recorded from different tissues.

Table 3.

Effects of NKCC1 inhibitors on slow waves_ICC in the rabbit and mouse small intestine

	Membrane potential, mV	Amplitude (upstroke), mV	Amplitude (plateau), mV	Frequency, min−1	Half-width, s	dV/dtmax, V/s	n
Rabbit
Control	−65 ± 3	53 ± 4	46 ± 2	9.8 ± 1.3	2.9 ± 0.4	9.5 ± 0.8	6
Bumetanide (10 μM)	−66 ± 2	52 ± 4	36 ± 1†	12.0 ± 0.9*	1.3 ± 0.2†	12.1 ± 1.9*	6
Control	−63 ± 3	56 ± 4	49 ± 3	9.3 ± 1.4	2.7 ± 0.3	11.4 ± 0.4	5
Furosemide (100 μM)	−63 ± 3	53 ± 5	38 ± 2†	11.0 ± 0.8*	1.4 ± 0.3†	11.8 ± 1.2	5
Mouse
Control	−68 ± 1	56 ± 4		23.9 ± 1.3	0.8 ± 0.1	2.0 ± 0.1	5
Bumetanide (10 μM)	−65 ± 2	49 ± 2*		24.5 ± 0.8	0.5 ± 0.1†	1.5 ± 0.1†	5
Control	−66 ± 1	53 ± 3		23.5 ± 0.9	0.9 ± 0.1	2.0 ± 0.1	5
Furosemide (100 μM)	−63 ± 2	44 ± 6*		23.3 ± 0.9	0.6 ± 0.1†	1.3 ± 0.2†	5

Values are means ± SD; n = number of animals.

^*P < 0.05,

^†P < 0.01, compared with Control.

The involvement of NKCC1 in the generation of the plateau component was further supported by the inhibitory effects of furosemide (100 μM), another NKCC1 inhibitor, on the plateau component of rabbit slow waves_ICC (Fig. 9B, Table 3). To evaluate the species-independent role of NKCC1 in the generation of the plateau component of slow waves_ICC, bumetanide (10 μM) and furosemide (100 μM) were applied to mouse slow waves_ICC. Both NKCC1 inhibitors decreased the plateau component of mouse slow waves_ICC (Fig. 9C, Table 3).

Fig. 9.

Effects of NKCC1 inhibitors on rabbit and mouse slow waves_ICC. A: rabbit slow waves_ICC were recorded before (Aa) and during application of 100 μM furosemide (Ab). B: expanded traces of rabbit slow waves_ICC recorded in the absence (Ba) and presence of 100 μM furosemide (Bb). C: mouse slow waves_ICC were recorded before (Ca) and during application of 10 μM bumetanide (Cb). D: expanded traces of mouse slow waves_ICC recorded in the absence (Da) and presence of 10 μM bumetanide (Db). The resting membrane potentials (dotted lines) were A, −67 mV; C, −67 mV.

NKCC1-immunoreactive cells in the myenteric region of the rabbit small intestine.

Immunohistochemical studies were performed to confirm expression of NKCC1 in rabbit ICC-MY. NKCC1-like immunoreactivity was observed in a network of cells in whole mount preparations of the longitudinal muscle layer (Fig. 10Aa). It should be noted that the myenteric plexus was also attached to the longitudinal muscle layer in these preparations of rabbit small intestine; however, this structure was not labeled in this study. The cells with NKCC1-like immunoreactivity were also immunoreactive for vimentin (Fig. 10, Ab and Ac) (28), which labels ICC in GI muscles (42). NKCC1 and vimentin antibodies colabeled the cell bodies and long processes of the network of cells (Fig. 10Ba). Double immunostaining for c-Kit and vimentin showed that the vimentin-positive cells are c-Kit-positive ICC (Fig. 10, Bb and Bc) (27, 29). These results suggest that ICC-MY is positive for Kit/vimentin/NKCC1 in the plane of the myenteric plexus of the rabbit small intestine.

Fig. 10.

NKCC1 expression in ICC-MY in the rabbit small intestine. Double immunostaining for Na⁺-K⁺-2Cl⁻ cotransporter (NKCC1, Aa) and the intermediate filament vimentin (Ab) in a whole mount preparation of the longitudinal muscle layer with the attached myenteric plexus of the rabbit small intestine. A merged image (Ac) indicated that the NKCC1-positive cellular network is also positive for vimentin. At higher magnification (Ba), NKCC1 immunoreactivity was observed in both the cell body (arrow) and long processes. Double immunostaining for c-Kit (Bb) and vimentin (Bc) revealed that the vimentin-positive cells (arrows) in the myenteric region of the rabbit small intestine are c-Kit-positive interstitial cells of Cajal. Scale bars: 50 μm (Aa–Ac), 20 μm (Ba–Bc).

DISCUSSION

Most studies to elucidate the pacemaker function of ICC distributed in the myenteric region (ICC-MY) of GI tract have used small laboratory animals (mouse, rat, or guinea pig). Electrophysiological evidence about pacemaking and membrane properties of ICC-MY in larger animals including humans is lacking. This might be due to technical difficulties in recording pacemaker potentials directly from ICC-MY in situ. In the present study, direct recording from ICC-MY of the rabbit small intestine was performed with intracellular microelectrodes to characterize the nature of slow waves_ICC. Rabbit slow waves_ICC had two phases, an upstroke component and a plateau component. Voltage-dependent Ca²⁺ influx appears to contribute to the upstroke component and Cl⁻ efflux through a Ca²⁺-activated Cl⁻ conductance contributes to the plateau component of rabbit slow waves_ICC. In addition, electrophysiological and immunohistochemical evidence suggests that NKCC1 plays an important role in the maintenance of the plateau component of slow waves_ICC.

It is well known that ICC generate slow waves (previously called “driving potentials in guinea pig gastric antrum” or “pacemaker potentials” in mouse small intestine) and these events conduct to smooth muscle cells (12, 20). Therefore, there is no need to call them by different terms (suggesting they are due to a different mechanism). The waveform delineates their source. This is analogous to the heart, where we don't call an action potential in the sinoatrial node as a pacemaker potential, in the atrium as a follower potential, and in the ventricle as an action potential. The events at all places are “action potentials” and the waveforms are tuned by local conductances as the events propagate through. Similar events occur in the GI tract: Slow waves are generated by ICC, and they conduct to circular and longitudinal smooth muscle cells. We think that the previous terminology is complex and may be confusing to researchers outside the field of ICC. Therefore, we would suggest the following terminology: Slow waves recorded from ICC are slow waves_ICC. Slow waves recorded from smooth muscle cells are slow waves_SMC (26, 38).

NiCl₂ or nominally Ca²⁺-free solution reduced the dV/dt_max and amplitude of the upstroke component of slow waves_ICC in both rabbit (the present study) and mouse small intestine (20). A recent study revealed that alpha_1H Ca²⁺ channel (also known as Ca_V3.2), an alpha subunit of T-type Ca²⁺ channel, contributes to the upstroke component of mouse slow waves_ICC in the small intestine (5, 50). However, this may not be the case with the upstroke component of rabbit slow waves_ICC, since the inhibitory effects of NiCl₂ or nominally Ca²⁺-free solution on the upstroke component were weaker in the rabbit slow waves_ICC compared with the effects on mouse slow waves_ICC. Moreover, replacement of Ca²⁺ with Sr²⁺ enhanced the activation of the upstroke component of rabbit slow waves_ICC but decreased those of mouse slow waves_ICC. These differences might be explained by different ion channels underlying the generation of the upstroke component of slow waves_ICC in the small intestine between rabbit and mouse. It has been reported that some of the T-type Ca²⁺ channels are resistant to the effects of Ni²⁺ (16). Thus it is possible that Ni²⁺-resistant T-type Ca²⁺ channels might be responsible for the upstroke component of rabbit slow waves_ICC. Recently, it was shown that Na⁺ channel (Na_V1.5) is expressed in ICC-MY in the human jejunum (39). Thus it is possible that the upstroke of rabbit slow waves_ICC may have contributions from additional voltage-dependent cation channels. However, it would be difficult to evaluate the role of extracellular Na⁺ in the generation of the upstroke component under the experimental conditions of the present study, because lowering extracellular Na⁺ concentration depolarized cells, resulting in the reduction of the amplitude and dV/dt_max of the upstroke component (Y. Kito, unpublished observations), as in the case of high K⁺-solution (18, 20). As described above, replacement of Ca²⁺ with Sr²⁺ increased the amplitude and dV/dt_max of the upstroke component of rabbit slow waves_ICC. This enhancement of the upstroke component could be induced by Sr²⁺ influx through voltage-dependent Ca²⁺ channels, since Sr²⁺ can carry charge through various types of voltage-dependent Ca²⁺ channels (1, 13, 33). Although the molecular identity of the ion channels responsible for the upstroke component of rabbit slow waves_ICC is unknown, the present results indicate that the upstroke component of rabbit slow waves_ICC seems to be supported by Ni²⁺-resistant, Sr²⁺-permeable, voltage-dependent Ca²⁺ channels. Although it is possible that T-type Ca²⁺ channels may carry the upstroke because cloned Ca_V3.1 channels are resistant to Ni²⁺ (16, 17), further investigation is needed regarding this issue.

The plateau component of rabbit slow waves_ICC was inhibited by low-[Cl⁻]_o solution, Cl⁻ channel blockers, and CPA, indicating that the plateau component is likely to be generated by the activation of CaCC, which is in agreement with previous reports in other slow waves_ICC (guinea pig gastric antrum, Ref. 18; mouse small intestine, Refs. 15, 20, 51). Application of low-[Cl⁻]_o solution transiently increased the amplitude of the plateau component of rabbit slow waves_ICC, and then this component decreased gradually (Fig. 5). These observations are consistent with an initial increase in driving force on Cl⁻ after the reduction in extracellular Cl⁻ and then gradual depletion in the Cl⁻ gradient. A transient increase in the amplitude of the plateau component of slow waves_ICC was also observed in guinea pig gastric slow waves_ICC but not in mouse small intestinal slow waves_ICC. Thus the contribution of Cl⁻ efflux during the plateau component of slow waves_ICC seems to be smaller in the mouse small intestine than that in the rabbit small intestine or guinea pig gastric antrum. The order of the duration of slow waves_ICC are as follows: guinea pig gastric antrum (7–10 s) (12, 18, 19) > rabbit small intestine (2–3 s) (the present study) > mouse small intestine (< 1 s) (20–23). Hence it is tempting to speculate that the longer duration of slow waves_ICC reflects the increased activity of CaCC. This would also mean that sustaining the openings of Ca²⁺-activated Cl⁻ channels is an important factor affecting the frequency of slow waves_ICC. Interestingly, inhibition of the plateau component of rabbit slow waves_ICC by DIDS, CPA, or bumetanide increased the frequency of slow waves_ICC. Similar results have been demonstrated in the guinea pig gastric antrum (18, 19) but not in the mouse small intestine (20). In contrast, inhibition of the upstroke component of rabbit slow waves_ICC by NiCl₂ decreased the frequency of slow waves_ICC. These findings demonstrate that the frequency of slow waves_ICC in the rabbit intestinal muscles is affected by the availability of ion channels responsible for the upstroke depolarization and by the duration of the plateau component.

NKCC1 transports extracellular Cl⁻ into the cell up its electrical chemical gradient, using energy stored in the Na⁺ gradient (4, 7, 10, 34, 49). Although costaining with typical markers of ICC (e.g., Kit and Ano1) was not possible in rabbit small intestine, and vimentin may not be ideally selective for ICC, NKCC1 overlapped with vimentin and vimentin overlapped with Kit in labeling studies of the rabbit intestine. These data support the idea that Kit and NKCC1 are coexpressed in the same cells. Furthermore, the morphology of the cells that were immunopositive for NKCC1 is the same as the cells identified as ICC by labeling with Kit antibodies, and electrophysiological recording showed that NKCC1 inhibitors reduce the plateau component of rabbit slow waves_ICC. Thus it seems plausible that NKCC1 inhibitors may reduce accumulation of Cl⁻ into ICC-MY, reducing Cl⁻ current via CaCC as the driving force for this conductance shifts toward more negative potentials. Since NKCC1 inhibitors also inhibited the plateau component of mouse slow waves_ICC, the role of NKCC1 in the maintenance of the plateau component of small intestinal slow waves_ICC seems to be species independent. Therefore, it is possible to formulate a hypothesis about Cl⁻ handling mechanisms in ICC-MY of the small intestine: Cl⁻ efflux occurs via CaCC and Cl⁻ recovery occurs via NKCC1 to maintain [Cl⁻]_i, which is essential to maintain the driving force for the plateau component of slow waves_ICC. Similar Cl⁻ dynamics were reported in olfactory sensory cilia, where odor stimulation causes Cl⁻ efflux through Ca²⁺-activated Cl⁻ channels (Ano2), leading to electrical excitation, followed by Cl⁻ accumulation into the ciliary lumen via NKCC1 or SLC4A1, Cl⁻/HCO₃⁻ exchanger, to maintain an elevated intracellular Cl⁻ concentration (10). It is worth noting that RMP of CSMCs was more depolarized in NKCC1 knockout mice compared with that in control mice (48). In the present study, NKCC1 inhibitors (100 μM bumetanide or 300 μM furosemide) transiently depolarized the membrane by ∼5 mV in rabbit ICC-MY (Y. Kito, unpublished observations). These observations imply that NKCC1 may also contribute to the RMP of ICC-MY.

In conclusion, electrophysiological properties of ICC-MY of the rabbit small intestine were studied in situ by use of intracellular recording. Slow waves recorded directly from ICC-MY (slow waves_ICC) had two components, an upstroke component followed by a plateau component. The upstroke component was generated, in part, by Ca²⁺ influx through voltage-dependent Ca²⁺ channels; the plateau component was dependent on Cl⁻ current through CaCC. The present findings also indicate that NKCC1 plays a critical role in regulating intracellular Cl⁻ homeostasis of small intestinal ICC-MY. The mechanisms underlying the generation of the upstroke component of slow waves_ICC in the small intestine appear to be different between rabbit and mouse.

GRANTS

This project was supported by the 24th General Assembly of the Japanese Association of Medical Sciences to Y. Kito. Funding for S. M. Ward and K. M. Sanders came from P01 DK41315 from the National Institute of Diabetes and Digestive and Kidney Diseases.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Y.K. conception and design of research; Y.K. and R.M. performed experiments; Y.K. and R.M. analyzed data; Y.K., R.M., S.M.W., and K.M.S. interpreted results of experiments; Y.K. and R.M. prepared figures; Y.K. and R.M. drafted manuscript; Y.K., R.M., S.M.W., and K.M.S. edited and revised manuscript; Y.K., R.M., S.M.W., and K.M.S. approved final version of manuscript.