Ca2+ influx through carbachol-activated non-selective cation channels in guinea-pig gastric myocytes

Authors


Corresponding author K. W. Kim: Department of Physiology and Biophysics, Seoul National University College of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul 110–799, Korea. Email : kimkw@plaza.snu.ac.kr

Abstract

  • 1Ca2+ microfluorometry (100 μm K5 fura-2) and the voltage-clamp technique were combined to study the effect of carbachol (CCh, 50 μm) in inducing currents (ICCh) through non-selective cation channels (NSCCCh) and increments in global cytosolic Ca2+ concentration (Δ[Ca2+]c).
  • 2In Na+-containing bath solution, ICCh fell from an initial phasic to a subsequent small (5 %) tonic component; Δ[Ca2+]c fell to zero. Tonic ICCh and [Ca2+]c became prominent after substitution of extracellular 140 mm Na+ by 140 mm Cs+. Tonic ICCh and Δ[Ca2+]c were insensitive to intracellular heparin (3 mg ml−1) and ryanodine (4 μm), i.e. they did not depend on Ca2+ release from sarcoplasmic reticulum (SR).
  • 3Single channel currents of NSCCCh could be resolved in whole-cell recordings. Substitution of Na+ by Cs+ increased NSCCCh activity by one order of magnitude and slope conductance from 22 to 30 pS. Extracellular quinidine (3 μm) reversibly blocked the NSCCCh activity.
  • 4Both tonic ICCh and tonic Δ[Ca2+]c (a) followed a similar time course of activation, desensitization and facilitation, (b) were reversibly blocked by 3 μm quinidine, and (c) persisted upon block of SR Ca2+ release.
  • 5A Ca2+ fractional current of tonic ICCh (fCa) of 0.009 was calculated by comparing the ratio Δ[Ca2+]c (corrected for simultaneous Ca2+ redistribution) over ICCh with depolarization-induced *Δ[Ca2+]c (Δ[Ca2+]c calculated from ICa induced by a 400 ms depolarization from −60 to 0 mV at 2 mm[Ca2+]o, 145 mm[Cs+]o) over ICa. fCa was 0.023 at [Ca2+]o= 4 mm.
  • 6With 110 mm extracellular CaCl2 and 145 mm intracellular CsCl, ICCh reversed at +19.5 mV suggesting a permeability ratio PCa/PCs of 2.8.
  • 7We conclude that Ca2+ influx through NSCCCh under physiological [Ca2+]o could induce Δ[Ca2+]c. The fCa was, however, much smaller than the one calculated from the reversal potential.

In smooth muscle cells (SMC), muscarinic agonists like acetylcholine or carbachol (CCh) activate contraction and Δ[Ca2+]c along a biphasic time course. Typically, a phasic Δ[Ca2+]c due to Ins P3-induced Ca2+ release (IICR) from the sarcoplasmic reticulum is followed by a slow tonic Δ[Ca2+]c attributed to influx of extracellular Ca2+ ions (Karaki et al. 1997). In intestinal SMC, the mechanisms of nifedipine-insensitive Ca2+ influx during the agonist-induced tonic Δ[Ca2+]c remain controversial (Putney, 1990; Irvine, 1992; Pacaud & Bolton, 1991; Wang et al. 1993; Sato et al. 1994; Komori et al. 1996).

In lymphocytes or secretory cells, tonic Δ[Ca2+]c may result from the ‘capacitative Ca2+ entry pathway’ that is activated when the Ins P3-sensitive Ca2+ pool is emptied (Putney, 1990; Ca2+-release activated Ca2+ channel (CRAC) current: Hoth & Penner, 1992; Parekh et al. 1997). In SMC, however, the presence of a CRAC current has rarely been proven (Wayman et al. 1998). Instead, tonic Δ[Ca2+]c has been attributed to Ca2+ influx through receptor-operated non-selective cation channels, for example the P2X-type ATP-gated receptor channel (Benham & Tsien, 1987; Benham, 1989; Schneider et al. 1991).

In intestinal SMC, muscarinic agonists activate an inward current (ICCh) through non-selective cation channels (NSCCCh) via a pertussis toxin-sensitive G protein (Inoue & Isenberg, 1990c). Although previous reports in intestinal myocytes have shown a close relationship between the amplitude of ICCh and Δ[Ca2+]c (Pacaud & Bolton, 1991; Komori et al. 1993), direct demonstration of Ca2+ influx through NSCCCh causing Δ[Ca2+]c is still missing. In extracellular 90 mm CaCl2 solutions, inward ICCh was measured, suggesting the existence of tonic Ca2+ influx, and from its reversal potential (ca+20 mV) a permeability ratio of Ca2+versus monovalent cations of ca 2 : 1 was extrapolated (Inoue & Isenberg, 1990a; Takai et al. 1997). It is questionable whether this permeability ratio would apply to Ca2+ influx from physiological Tyrode solution, where influx of monovalent cations may alter the permeation of Ca2+ ions in the pore (e.g. Aidley & Standfield, 1996).

Another estimate of Ca2+ influx compares the agonist-induced inward current and tonic Δ[Ca2+]c with the depolarization-induced ICa and Ca2+ transient (*Δ[Ca2+]c), the latter caused by Ca2+ influx through highly selective Ca2+ channels (Schneider et al. 1991). In neuronal cells, this estimate of the ‘Ca2+ fractional current’ (fCa) was improved by the method of ‘added-buffer approach’ (see Neher, 1995).

In neurons, the episodes of Ca2+ influx with non-selective cationic currents were usually short and fast, hence one could neglect the effects of the slow Ca2+ redistribution processes such as sequestration into the SR or pumping across the plasma membrane (Neher, 1995). In the present experiments with gastric SMC, however, tonic ICCh activated along a very slow time course and Ca2+ redistribution effects had to be taken into account (see Guerrero et al. 1994). Our results show that Ca2+ ions carry ca 1 % of tonic ICCh when the extracellular solution contains 2 mm Ca2+ and 135 mm Cs+. This measured value of fCa is much smaller than the *fCa (fCa calculated from the reversal potential under bi-ionic conditions, Schneggenburger et al. 1993) of ca 10 % previously extrapolated from the ICCh reversal potential measured in 90 mm CaCl2 extracellular solution.

methods

Cell dissociation

Gastric myocytes were isolated enzymatically from the antral circular layer of guinea-pig stomach as described previously (Kim et al. 1995). Briefly, guinea-pigs of either sex weighing 300–350 g were exsanguinated after stunning. The antral part of the stomach was cut and the circular muscle layer was dissected from the longitudinal layer using fine scissors and then cut into small segments (2–3 mm). The tissue chunks were then incubated for 20–25 min at 35°C in a digestion medium consisting of a Ca2+-free Tyrode solution (see below) containing 0.15 % collagenase (Wako or Sigma Type IA), 0.05 % dithiothreitol, 0.1 % trypsin inhibitor and 0.2 % bovine serum albumin. Single myocytes were dispersed by gentle agitation of the digested segments with a wide-bore glass pipette. Isolated myocytes were kept at 4°C until use. All experiments were carried out within 12 h of harvesting cells and the bath solution was warmed by a circulating water jacket just before perfusing the experimental bath, the volume of which was about 0.1 ml. The temperature of the bath solution was monitored and maintained at 28 ± 1°C.

Electrophysiological recordings

Whole-cell membrane currents were measured with an Axopatch-1D patch-clamp amplifier (Axon Instruments) filtered at 5 kHz. Glass pipettes with a resistance of 2–3 MΩ were used to make a gigaseal. pCLAMP v.5.7.2, Axoscope v.1.0 and Digidata-1200 (all from Axon Instruments) were used for the acquisition of data and applying command pulses. For the continuous recording of ICCh, the data were sampled at 10 or 200 Hz using Axoscope and displayed on a computer monitor. The recording of voltage-operated Ca2+ current for fractional Ca2+ current (Fig. 5) was done with pCLAMP and sampled at 1 kHz. Data were analysed using pCLAMP and Origin (Microcal Software Inc., USA).

Figure 5.

Ca2+ dependence of cytoplasmic Ca2+ removal processes

A, ICa induced by a 400 ms depolarization from −60 to 0 mV. Note that the time scale is faster than those of following figures. B, *Δ[Ca2+]c caused by ICa. Decay of *Δ[Ca2+]c was fitted by a single-exponential function (smooth dashed line). C, the exponential function in B was differentiated to obtain the time course of the Ca2+ removal process (smooth dashed line). D, rate of Ca2+ removal plotted against the corresponding global *[Ca2+]c (▪). Differentiated results of raw data were plotted together in C (noisy line) and D (scattered open circles). Note that the removal rate of [Ca2+]c depends linearly on [Ca2+]c (straight line). Heparin (4 mg ml−1) and ryanodine (4 μm) were dialysed from the pipette.

Openings of single channels were discernible under magnified whole-cell recording conditions as described by Inoue et al. (1987). In the whole-cell configuration, long-lasting application of 100 μm CCh induced a marked desensitization of ICCh, the low channel activity permitting single channel recordings. Unitary currents were amplified and filtered at 1 kHz with an Axopatch 1-D amplifier. The data were stored on a digital tape recorder (DTR 1204, Biologic, France) for later analysis. Recorded data were played back and digitized using Digidata 1200 at 1 kHz and low-pass filtered at 300 Hz for illustration. Results were analysed by using pCLAMP 6.02 and Origin software. Single channel amplitudes and the open state probability of N channels (NPo) were calculated from amplitude histograms. NPo values of the data recorded at steady membrane potential (−60 mV) were determined from:

display math(1)

where the tj is the time spent with j= 1, 2 …N channels open, N is the number of channels and T is the duration of the measurement.

Fluorescence measurements

[Ca2+]c was measured with a microfluorimeter consisting of an inverted fluorescence microscope (Diaphot 300, Nikon) with a × 40, 0.85 NA objective, a photomultiplier tube (type R 1527, Hamamatsu, Japan) and PTI deltascan illuminator (monochromator system with 2 nm bandpath; Photon Technology International Inc., USA). Light was provided by a 75 W xenon lamp (Ushino, Japan) through a chopper wheel (frequency of 4 or 10 Hz) which alternated the light path to monochromators (340 and 380 nm, respectively). A 425 nm short-pass excitation filter reduced background fluorescence. A 570 nm short-pass dichroic mirror passed emission light onto the photomultiplier. A mechanical diaphragm situated at an image plane in the emission path limited the measurement to a single cell.

Fura-2 loading and estimation of [Ca2+]c

K5 fura-2 was dissolved in distilled water to give a 10 mm stock solution and diluted by pipette solution to a final concentration of 100 μm. After the whole-cell configuration was achieved, the myocyte was dialysed for 4–5 min before the start of the experiment. Within this time, the absolute fluorescence of dialysed fura-2 usually reached a steady-state value. [Ca2+]c was calculated from the following equation (Grynkiewicz et al. 1985):

display math(2)

where Kd is the effective dissociation constant of fura-2, and b is the ratio of fluorescence signals at 380 nm (Sf2/Sb2) without Ca2+ (Sf2) and with saturating Ca2+ (Sb2). Rmin represents a ratio of 340/380 in the absence of Ca2+, and Rmax represents this ratio when Ca2+ concentration is at saturation point. Rmin was obtained by dialysing the cells with 10 mm BAPTA-containing solution (Rmin= 0.4 ± 0.01, n= 4). Rmax was obtained by stepping the membrane potential to approximately −200 to −250 mV, which caused membrane leakage and induced a large increase in R. Two minutes after the hyperpolarizing step R had stabilized, this value was taken as Rmax. In cells where the measurement of Rmax was missed, the previously obtained mean value (8.6 ± 0.26, n= 15) was used. The value of Kdb (1.135 μm) was estimated by perfusing the cells with a pipette solution containing 10 mm BAPTA-buffered solution, where [Ca2+] was clamped at 100 nm by adding 2.97 mm CaCl2 according to a computer program (Schoenmakers et al. 1992). Background fluorescence was determined in the cell-attached configuration, and was subtracted from the respective wavelength.

Solutions and drugs

The physiological salt solution (Na+ Tyrode solution) contained (mm): NaCl, 135; KCl, 5; CaCl2, 2; MgCl2, 1; glucose, 5; Hepes, 10; and pH was adjusted to 7.4 with NaOH. In most experiments recording ICCh, both NaCl and KCl were replaced with the same concentration of CsCl (total 140 mm) and the pH was adjusted to 7.4 with CsOH (Cs+ Tyrode solution, total Cs+ was approximately 145 mm). The pipette solution consisted of (mm): CsCl, 140; Na2ATP, 2.5; Tris GTP, 0.2; MgCl2, 4; Hepes, 10; K5 fura-2, 0.1; pH was adjusted to 7.3 with CsOH. For comparison, some experiments were performed with KCl and NaCl instead of CsCl in the intra- and extracellular solution (Fig. 1). When the external cations were isosmotically replaced by N-methyl-D-glucamate (NMDG+) or Ca2+, a 3 m KCl agar bridge reference electrode was used, and corrections were made for liquid junction potentials but not for ionic activities and surface charge screening effect of Ca2+. To apply drugs, the experimental chamber was superfused by gravity at a rate of 2 ml min−1. K5 fura-2 was purchased from Molecular Probes; all other drugs were purchased from Sigma.

Figure 1.

Whole-cell currents (ICCh) and Δ[Ca2+]c induced by 50 μm CCh modified by substitution of extracellular Na+ by Cs+

A, in Na+ Tyrode solution, 50 μm CCh induced phasic responses, i.e. both ICCh and Δ[Ca2+]c rapidly decayed. Substitution of Na+ by Cs+ recovered a tonic ICCh and a tonic Δ[Ca2+]c. KCl solution was in the electrode and 1 μm verapamil was in the bath solution to block L-type Ca2+ channels. Top trace: holding potential −40 mV, 2 s pulses to 0 mV at 0.1 Hz. B, dialysis of Cs+ electrode solution containing 3 mg ml−1 heparin blocks the phasic response. CCh induces a tonic ICCh and a tonic Δ[Ca2+]c that rise along a slow time course (holding potential −60 mV). C, Cs+ electrode solution containing 4 μm ryanodine and 3 mg ml−1 heparin. Lack of response to 10 mm caffeine suggests that SR has been functionally removed; nevertheless CCh induced tonic ICCh and tonic Δ[Ca2+]c of the usual amplitude (holding potential −60 mV).

Data are presented as the mean ±s.e.m. Significant differences were detected using Student's unpaired t test (P < 0.05).

results

The phasic and tonic component of ICCh and Δ[Ca2+]c

Figure 1A illustrates the effects of CCh on membrane current and [Ca2+]c typical for a cell superfused with a normal Na+ Tyrode solution containing 135 mm NaCl, 5 mm KCl and 2 mm CaCl2 as well as 1 μm verapamil to block interference from Ca2+ influx through L-type Ca2+ channels. The electrode was filled with a KCl solution. At a holding potential of −40 mV, the membrane current (upper trace of Fig. 1A) was approximately zero and the [Ca2+]c was ca 180 nm (lower trace of Fig. 1A). Bath application of 50 μm CCh induced an inward current (ICCh) that reached a peak of −430 pA and fell within the following 40 s to a small ‘tonic component’ of −20 pA. The difference peak ICCh minus tonic ICCh defines the ‘phasic component’ of ICCh. The phasic component of ICCh had its counterpart in a phasic Δ[Ca2+]c that peaked within 5 s to 550 nm. From the peak, Δ[Ca2+]c fell within 60 s towards the basal 180 nm level, i.e. there was no prominent tonic Δ[Ca2+]c component in cells superfused with Na+ Tyrode solution.

The global increase in [Ca2+]c should have a counterpart in the subsarcolemmal Δ[Ca2+] as has been reported by the increase in Ca2+-activated IK(Ca) (e.g. Ganitkevich & Isenberg, 1995). Superimposition with ICCh was avoided by analysing IK(Ca) during clamp steps to 0 mV, the reversal potential of ICCh (downward arrows, top trace of Fig. 1A). However, CCh increased IK(Ca) only slightly and the effect was observed in only 2 of 5 cells (Fig. 1A). CCh increased IK(Ca) by an amount that was much smaller than expected from Δ[Ca2+]c of 200 nm. A speculative interpretation could be that CCh depleted releasable Ca2+ thereby preventing spontaneous Ca2+ release and spontaneous transient outward currents (STOCs). Alternatively it might be possible that CCh suppresses the Ca2+-activated maxi-K+ channel (Cole et al. 1989).

In the continued presence of 50 μm CCh, the substitution of extracellular Na+ by Cs+ largely augmented the small tonic ICCh component (Fig. 1A: from −20 to −440 pA). The substitution induced a tonic Δ[Ca2+]c, i.e. with an ∼5 s time delay [Ca2+]c increased from basal level (180 mm) to 350 nm. The existence of tonic ICCh and Δ[Ca2+]c in Cs+ Tyrode solution provided the possibility to analyse the relation between Ca2+ influx via tonic ICCh and resulting Δ[Ca2+]c. Hence, all following data were obtained with Cs+ Tyrode bath solution. Also the pipette solution was changed to a CsCl solution in order to block K+ currents superimposed on ICCh. The superimposition of voltage-operated Ca2+ influx was excluded by analysing the relation at a holding potential of −60 mV.

The phasic components of ICCh and Δ[Ca2+]c are due to IICR

The phasic component of ICCh (Fig. 1A) was thought to be a consequence of InsP3-induced Ca2+ release (IICR) and phasic Δ[Ca2+]c since elevated [Ca2+]c has been reported to facilitate ICCh (Inoue & Isenberg, 1990b; Pacaud & Bolton, 1991; Komori et al. 1993). Heparin prevents agonist-induced IICR in smooth muscle cells (Kobayashi et al. 1988). Hence, 3 mg heparin was added per millilitre of CsCl electrode solution dialysing the cell. The presence of intracellular heparin suppressed the phasic components of ICCh and Δ[Ca2+]c; CCh induced a slow rise of both signals which achieved their maximum within 30 s. Normalized to this maximum, ICCh and Δ[Ca2+]c had relative amplitudes of 0.49 ± 0.053 and 0.22 ± 0.047, respectively, at 10 s after CCh application (n= 16). At 20 s, the relative amplitudes were 0.90 ± 0.025 and 0.68 ± 0.042, respectively.

Ca2+-induced calcium release (CICR) does not contribute to CCh-induced tonic Δ[Ca2+]c

Ca2+-induced Ca2+ release (CIRC) through sarcoplasmic reticulum (SR) ryanodine receptors (RyRs) may have been activated by Ca2+ influx and local increments in [Ca2+]c. To define a possible contribution of CICR to the CCh-induced tonic Δ[Ca2+]c, the CCh effects on Δ[Ca2+]c were reanalysed with CsCl electrode solutions including both ryanodine (4 μm) and heparin (3 mg ml−1). Since the ryanodine block of RyRs is use dependent (Rousseau et al. 1987), the experiments started with a caffeine application (10 mm) that opens RyRs. In six cells dialysed with ryanodine, the first caffeine application evoked no or only a small Δ[Ca2+]c (Fig. 1C). In four other cells dialysed with ryanodine, the first caffeine application induced a transient Δ[Ca2+]c. To these cells, we applied caffeine again and confirmed the absence of a caffeine-induced Δ[Ca2+]c, which indicated that CICR was non-functional in this condition. Nevertheless, the subsequent application of 50 μm CCh induced the tonic ICCh and Δ[Ca2+]c as in the absence of ryanodine (Fig. 1C).

Extracellular Cs+ facilitates tonic ICCh and Δ[Ca2+]c by increasing NPo

As mentioned above, CCh induced a large tonic ICCh with Cs+ but not with Na+ as the main charge carrier (compare Zholos & Bolton, 1995). In the continued presence of CCh, the ratio of tonic ICCh carried by Cs+ over tonic ICCh carried by Na+ was 14.7 ± 1.45 (n= 14). Since the voltage-dependent gating of ICCh is insensitive to the substitution of Na+ by Cs+ (Zholos & Bolton, 1995), we decided to study the effect of external cation replacement on open probability (Po) and conductance of the single CCh-activated non-selective cation channels (NSCCCh). Single channel activities could be resolved in whole-cell recordings after largely suppressing the gross inward current by repeated CCh applications leading to desensitization (see Methods). Figure 2A shows single channels from the same myocyte, activated by 2 μm CCh in Cs+ Tyrode (Fig. 2A) and by 15 μm CCh in Na+ Tyrode solution (Fig. 2B). Despite the simultaneous increase in agonist concentration, the substitution of Cs+ by Na+ reduced the single channel activity NPo from 0.75 to 0.10.

Figure 2.

Currents through single CCh-activated non-selective cation channels (NSCCCh) recorded under whole-cell configuration

A and B, after severe desensitization of ICCh (several long-lasting CCh applications) single channel currents are discernible on a magnified current scale (membrane potential −60 mV). In the same cell, 2 μm CCh was applied in Cs+ Tyrode solution (A) and 15 μm CCh was applied in Na+ Tyrode solution (B). Ab and Bb show NSCCCh recordings on an expanded time scale. C, amplitude histograms of NSCCCh for 15 s sampling periods and least square fit by a sum of Gaussian distributions in Cs+ Tyrode solution (Ca) and Na+ Tyrode solution (Cb). Cc, I–V curves of open channel current recorded in Cs+ Tyrode solution (▪) and Na+ Tyrode solution (•). Extrapolated lines crossed the horizontal axis at −1.6 and −0.5 mV in Cs+ and Na+ Tyrode solution, respectively. D, reversible block of NSCCCh by bath-applied quinidine (3 μm) (different cell).

The amplitudes of the single channel current were evaluated with histograms (Fig. 2Ca and b). In Cs+ Tyrode solution, the simultaneous opening of four channels generated four peaks (Fig. 2Ca: from right to left separated by 1.7, 1.6, 1.4 and 1.6 pA). In Na+ Tyrode solution, however, the histogram showed only a single peak at 1.3 pA (Fig. 2Cb). The plot of single channel current amplitude over the membrane potential yielded a unitary slope conductance of 29 pS in the case of Cs+ and of 21 pS in the case of Na+ Tyrode solution, respectively (Fig. 2Cc). Mean unitary conductances of four cells analysed were 30.0 ± 0.24 and 21.8 ± 0.60 pA in Cs+ and Na+ Tyrode solution, respectively. Both plots converged to a reversal potential of 0 mV, as can be expected for a non-selective cation-permeable channel (i.e. NSCCCh).

In another three cells, the blocking effect of quinidine on ICCh (see next paragraph) was tested on the single channel level. Quinidine (3 μm) effectively and reversibly blocked the NSCCCh activity (Fig. 2D).

Quinidine block of CCh-induced tonic ICCh and Δ[Ca2+]c

The above results suggested that the tonic Δ[Ca2+]c may be caused by Ca2+ influx through NSCCCh. Since the openings of NSCCCh were blocked by bath-applied quinidine (Fig. 2D), the effect of quinidine was tested on CCh-induced tonic ICCh and Δ[Ca2+]c (heparin in Cs+ pipette solution). Quinidine (3 μm) reduced ICCh by 99 ± 0.6 % (n= 7) and tonic Δ[Ca2+]c by 96 ± 1.3 %. Quinidine at 0.5 μm inhibited ICCh and Δ[Ca2+]c by 52 ± 11.4 % and 46 ± 9.5 %, respectively (n= 6). Figure 3A shows that the block was reversible upon washout of quinidine. Although not shown here, dialysis of 0.3 mm GTPγS from the pipette solution induced a slowly increasing inward current as well as a Δ[Ca2+]c with similar time course. Bath application of 3 μm quinidine reversibly blocked the GTPγ-induced inward current and Δ[Ca2+]c (n= 5).

Figure 3.

Suppression of ICCh and tonic Δ[Ca2+]c by quinidine (top) or by changing membrane potential (bottom)

A, 3 μm quinidine reversibly blocked CCh-induced ICCh and Δ[Ca2+]c along a parallel time course (symmetrical CsCl solutions, heparin present). B, the first control CCh response at −40 mV was followed by a 2 min washout, then CCh was applied a second time. In the continued presence of CCh the clamp potential was changed as indicated in the upper trace. Note: amplitude of tonic Δ[Ca2+]c (lower trace) followed the amplitude of tonic ICCh and not the Ca2+ driving force.

Dependence of tonic ICCh and tonic Δ[Ca2+]c on membrane potential

It is known that the activity of NSCCCh is suppressed by hyperpolarizing the membrane potential (Inoue & Isenberg, 1990a; Kim et al. 1995). Hence, we tested whether the CCh-induced tonic Δ[Ca2+]c follows the voltage dependence of ICCh. With symmetrical Cs+ solutions, a clamp step to the reversal potential of ICCh (0 mV) abolished the ICCh and induced a decay of tonic Δ[Ca2+]c, which was reversed by repolarization to −40 mV (Fig. 3B). Hyperpolarization to −60, −80 and −100 mV further increased tonic ICCh, however, with a sub-linear voltage dependence (Fig. 3B). Similarly, the hyperpolarizing clamp steps did not linearly increase Δ[Ca2+]c as expected from the increments in the Ca2+ driving force (Fig. 3B, see the step to −100 mV).

Desensitization and Ca2+ facilitation of the CCh responses

Upon repeated CCh applications, ICCh decreased in amplitude due to ‘desensitization’ (Inoue & Isenberg, 1990b). Desensitization can partially be compensated by the facilitating effects of Ca2+ influx through L-type Ca2+ channels (Ca2+-induced facilitation of ICCh, Inoue & Isenberg, 1990b; see also Fig. 1 of Kim et al. 1998). Here, we tested whether the superimposition of Ca2+ influx through L-type Ca2+ channels is able to facilitate ICCh and tonic Δ[Ca2+]c.

Upon the fourth CCh application both ICCh and Δ[Ca2+]c increased along a very slow time course (start at arrows labelled 4 in Fig. 4) because the preceding CCh exposures had induced desensitization (second and third CCh application not shown in Fig. 4). The simultaneous activation of L-type ICa by an 800 ms clamp step (from −60 to 0 mV) induced a sharp rise in [Ca2+]c. However, the inward current continued as tonic ICCh as long as CCh was present (approx. 40 s). The tonic ICCh was paralleled by a tonic Δ[Ca2+]c. The tonic elevation of [Ca2+]c contrasts with the rapid decay of *Δ[Ca2+]c due to ICa before and after the fourth CCh application. Figure 4 indicates this potentiating effect of CCh on the ICa-induced *Δ[Ca2+]c by comparison of the second with the first and third *Δ[Ca2+]c. Facilitation of CCh-induced ICCh and Δ[Ca2+]c by ICa was observed in five cells altogether. For the experiment in Fig. 4, ICa causing *Δ[Ca2+]c was augmented by 0.3 μm Bay K 8644. The traces at the top of Fig. 4 indicate during the second pulse a smaller amplitude for ICa than for ICa during the first pulse, an effect that could be due to a inhibition of ICa by the muscarinic stimulation in smooth muscle (Unno et al. 1995).

Figure 4.

Desensitization and Ca2+ facilitation of the CCh response. ICCh and Δ[Ca2+]c are compared with depolarization-induced ICa and *Δ[Ca2+]c

The cell was exposed to 50 μm CCh four times: ICCh and Δ[Ca2+]c during the first and fourth exposure are shown. Because of desensitization, the fourth response started at a very slow rate (start marked by arrows labelled 4). Before, during and after the fourth CCh application, 800 ms clamp steps from −60 to 0 mV induced ICa and *Δ[Ca2+]c (marked by asterisks). Uppermost traces show ICa on an expanded time scale. Note: the small and slow fourth CCh response was facilitated by superimposed ICa and spiky *Δ[Ca2+]c. Note: *Δ[Ca2+]c due to ICa decayed rapidly in contrast to superimposed tonic ICCh and tonic Δ[Ca2+]c that lasted as long as CCh was present. Symmetrical CsCl conditions, 0.3 μm Bay K 8644 present in the bath.

Calculation of fCa, the fraction of ICCh carried by Ca2+

Figure 5 analyses the *Δ[Ca2+]c that is caused by a depolarization activating Ca2+ influx through L-type Ca2+ channels. In comparison with the CCh-induced Δ[Ca2+]c, ICa-induced *Δ[Ca2+]c increased at a faster rate and reached a higher peak. We attribute this difference to the different Ca2+ selectivity of L-type Ca2+ channels versus the non-selective cation channels. While L-type ICa is carried nearly exclusively by Ca2+ ions (PCs/PCa= 0.0002, see Table 5.1 in Aidley & Stanfield, 1996), ICCh is carried mostly by Cs+ ions and only a small fraction is carried by Ca2+ ions. In the following section, the fraction of ICCh carried by Ca2+ ions (fractional Ca2+ current, fCa) will be evaluated from the comparison of *Δ[Ca2+]c and ICa with tonic Δ[Ca2+]c and ICCh, all four signals being measured in the same cell. Possible contributions of IICR and CICR were blocked by intracellular dialysis of heparin and ryanodine. However, because of the slow time course of tonic ICCh and Δ[Ca2+]c, the effect of simultaneous Ca2+ redistribution processes had to be incorporated (see Guerrero et al. 1994).

The CCh-induced Δ[Ca2+]c should relate to the time integral of ICCh according to:

display math(3)

where Δ[Ca2+]c(t) is the measured CCh-induced increase in free Ca2+ concentration over baseline at time t after CCh application. On the right, the term (fCa/(zBF Vol)) ΣICChΔt represents an ‘uncorrected’ fictitious increment in [Ca2+]c due to Ca2+ influx through NSCCCh, which will be reduced by simultaneous Ca2+ removal processes (ΣRCaΔt). z, B, F and Vol indicate the equivalence charge of Ca2+ (+2), the intracellular Ca2+ buffering power (Kamishima & McCarron, 1996), the Faraday constant, and the volume of cytosol in which the Ca2+ ions distribute, respectively. A similar expression holds true for the increment in Ca2+ concentration due to ICa:

display math(4)

where the time integral (∫ICa dt) yields the charge carried by ICa during the 400 ms of the depolarizing clamp step and where *Δ[Ca2+]c is the amplitude of measured peak minus basal [Ca2+]c (e.g. 660 nm in the experiment of Fig. 5B). In the context of eqn (4) the Ca2+ removal processes were neglected because they are slow in comparison with the rate of Ca2+ influx with ICa. Since the experiment was performed in the same cell, the available cell volume is identical. The monoexponential decay of *[Ca2+]c (Fig. 5B) suggests that the Ca2+ buffering capacity is unlikely to be exhausted, hence the B values in eqns (3) and (4) should be similar. ICa and *Δ[Ca2+]c can be measured directly and exactly, hence substitution and rearrangement of eqns (3) and (4) leads to eqn (5) where the cell volume (Vol) and the buffering power (B) are no longer included:

display math(5)

Equation (5) indicates that the plot of (Δ[Ca2+]c(t) +ΣRCaΔt) versus (*Δ[Ca2+]c/∫ICadt) ΣICChΔt has a slope fCa which is the fractional Ca2+ current of ICCh. The data necessary for the plot were measured in the same cell, starting with ICa and Δ[Ca2+]c and CCh-induced ICCh and Δ[Ca2+]c after an interval of 2 min (Fig. 6). The Ca2+ removal rate (RCa) was estimated from the decay of *Δ[Ca2+]c at −60 mV. In all experiments, the decline of *Δ[Ca2+]c could be fitted by a single exponential, and the rate of Ca2+ removal was a linear function of [Ca2+]c (Fig. 5D). This relationship (RCaversus ([Ca2+]c)) was applied to the CCh-induced Δ[Ca2+]c (Fig. 6A), estimating RCa at discrete steps of 1 s. The sum of removed [Ca2+]c over time (ΣRCaΔt) was added to the measured [Ca2+]c (left term of eqn (5), Fig. 6Ba). In Fig. 6Bb, (*Δ[Ca2+]c/∫ICadt) ΣICChΔt was plotted over the time of CCh exposure using discrete steps of 1 s. Finally, the fractional Ca2+ current, fCa, was evaluated according to eqn (5) from the slope of (Δ[Ca2+]cRCaΔt) (Fig. 6Bc, accum. [Ca2+]c fura-2) versus (*Δ[Ca2+]c/∫ICadt) ΣICChΔt (Fig. 6Bc, accum. [Ca2+]c current). The particular experiment illustrated in Fig. 6 yielded the result that 1 % of ICCh is carried by Ca2+ ions, i.e. fCa was 0.01.

Figure 6.

Evaluation of fractional Ca2+ current, fCa, of ICCh

A, CCh-induced ICCh and Δ[Ca2+]c (same myocyte as in Fig. 5). B, to evaluate fCa, data were transformed and plotted according to eqn (5): Δ[Ca2+]c(t) +ΣRCaΔt=fCa (*Δ[Ca2+]c/∫ICadt) ΣICChΔt. Ba, ordinate Δ[Ca2+]c(t) +ΣRCaΔt: removal-corrected accumulation of fura-2-measured [Ca2+]c plus Δ[Ca2+]c(t) is plotted against experimental time. Bb, CCh-induced accumulated Σ[Ca2+]c(t) was calculated from ΣICChΔt by multiplication with the ratio *Δ[Ca2+]c/∫ICadt (from Fig. 5) and plotted against experimental time. Note: in Ba and Bb CCh application started 3 s after zero time. Bc, plot according to eqn (5). Removal-corrected measured [Ca2+]c(t) (Ba) was plotted versus calculated Σ[Ca2+]c(t) (Bb) for 1 s time intervals. Linear regression yielded a slope of 0.0098 which identified fCa for this cell.

The calculation of fCa was performed in a total of eighteen cells for a membrane potential of −60 mV. On average, fCa was 0.009 ± 0.001. fCa was also calculated from experiments where only heparin but no ryanodine was present in the pipette solution. The resulting fCa value of 0.009 ± 0.0017 (n= 11) was not significantly different from the value obtained in the presence of ryanodine, suggesting that the contribution of CICR to CCh-activated Δ[Ca2+]c is insignificant. In six additional experiments, the extracellular [Ca2+]o was doubled from 2 to 4 mm; under these conditions fCa was 2.3 ± 0.32 % (−60 mV).

Electrophysiological estimation of the Ca2+ permeability ratio

Under bi-ionic conditions, the ratio of Ca2+ over Cs+ permeability through NSCCCh (PCa/PCs) can be calculated from the reversal potential (Vrev) of ICCh. Applying the Goldman-Hodgkin-Katz voltage equations, one obtains eqn (6) where R is the gas constant and T is the absolute temperature (compare Fieber & Adams, 1991; Zeilhofer et al. 1996):

display math(6)

Experimentally, Cs+ Tyrode solution was replaced with 110 mm CaCl2 bath solution and membrane potential was held at −60 mV. For this experiment, the CsCl pipette solution contained 2 mm EGTA pipette but no fura-2. CCh induced an inward current that decayed with time (Fig. 7A). Ramp-like pulses (0.11 V s−1, see inset) were applied before and during the stimulation with CCh. The CCh-induced difference current during the hyperpolarization was obtained by digital subtraction (Fig. 7B). The reversal potential was 19.5 ± 0.77 mV (n= 11), a value similar to the one obtained for the ACh-activated non-selective cationic current in other smooth muscle cells (Inoue & Isenberg, 1991a; Takai et al. 1997). According to eqn (6), Vrev= 20 mV translates into a permeability ratio of 2 : 1. Considering that the activity coefficients are smaller in 110 mm CaCl2 (0.52) than 145 mm CsCl solution (0.72), the permeability ratio PCa/PCs would be 2.9. Assuming a constant electrical field across the cell membrane, one can extrapolate fCa for the usual [Ca2+]o= 2 mm, [Cs+]c= 145 mm and holding membrane potential (Vhold)=−60 mV (see Schneggenburger et al. 1993):

display math(7)
Figure 7.

Voltage dependence and reversal potential of ICCh in 110 mm CaCl2 bathing solution

A, ramp-like repolarizations (0.11 V s−1, see inset) were applied before and during stimulation with CCh (50 μm). B, difference current during hyperpolarization was plotted as a function of membrane potential.

Using the above values, we calculate a *fCa= 0.14 that is more than tenfold larger than the fCa obtained from eqn (5).

discussion

The present study on the CCh response of gastric myocytes separated a phasic component due to IICR from a tonic component due to Ca2+ influx. Tonic ICCh and tonic Δ[Ca2+]c were highly correlated: both signals persisted after block of IICR and CICR, followed a similar time course during activation and spontaneous decay (desensitization), and were reversibly blocked by 3 μm quinidine. Also, Δ[Ca2+]c fell together with ICCh when ICCh was reduced by changes in the membrane potential. In summary, these results support the conclusion that Ca2+ influx with ICCh is the cause of tonic Δ[Ca2+]c.

In a variety of different cell types, agonists have been reported to induce Δ[Ca2+]c via Ca2+ influx through ICRAC caused by IICR emptying intracellular Ca2+ stores (see Introduction). In the present study on tonic CCh-induced Δ[Ca2+]c in the gastric SMC, a significant contribution of ICRAC is unlikely because Ca2+ release (necessary for activation of ICRAC) was blocked by heparin and ryanodine. Further, the properties of tonic ICCh (present study) differ from those of ICRAC (see e.g. Yao & Tsien, 1997). The reversal potential of ICCh is close to zero (present study), that of ICRAC is +40 mV. Only ∼1 %ICCh is carried by Ca2+ ions whilst ICRAC is a highly selective Ca2+ current. In summary, the present results are incompatible with the idea that ICCh is an ICRAC; instead, they support the idea that ICCh flows through non-selective cation channels.

f Ca= 0.01: possible sources of error

The present results from gastric myocytes indicate that ICCh can cause a substantial tonic Δ[Ca2+]c, despite the result that only 1 % of this current is carried by Ca2+ and 99 % is carried by Cs+ ions. This fractional Ca2+ current increased to 2 % when [Ca2+]o was doubled from 2 to 4 mm. The fractional Ca2+ current was obtained by comparing Ca2+ transients and inward currents mediated by ICCh and ICa. Measuring the signals in the same cell avoided the problem of estimating the buffer value or the cytosolic distribution volume of the Ca2+ ions. Nevertheless, errors may have been introduced by the variety of necessary other assumptions. Errors due to the concomitant Ca2+ removal were corrected (according to Guerrero et al. 1994). The assumption that CCh could have induced Ca2+ influx via other mechanisms such as ICRAC has already been discussed as unlikely. The analysis of fCa along Fig. 6 assumes that the Ca2+ influx would not exhaust the endogenous Ca2+ buffers, i.e. that the buffer value B would be constant; indeed a constant slope for the ratio *Δ[Ca2+]c/∫ICadt has been measured recently in gastric myocytes (Kim et al. 1997). The assumption that CCh does not modulate the rate of Ca2+ removal, via Ca2+ ATPases and Na+-Ca2+ exchange, unfortunately cannot be tested by the present experiments. The final assumption that neither IICR nor CICR would contribute to the Δ[Ca2+]c signals was validated by blocking IICR and CICR with heparin and ryanodine, respectively. Additional experiments without ryanodine resulted in identical fCa values suggesting the relative unimportance of CICR even for *Δ[Ca2+]c induced by ICa for the ‘calibration’ procedure (Fig. 7; see also Kim et al. 1997).

*fCa extrapolated from Vrev is unlikely to be correct

The reversal potential (Vrev) of ICCh suggested that the CCh-activated non-selective cation channel is twice as permeable for Ca2+ than for Cs+ ions (see also Takai et al. 1997), similar to the value reported for the P2x-gated non-selective cation channels (Benham & Tsien, 1987; Benham, 1989; Fieber & Adams, 1991; Schneider et al. 1991). When we extrapolated this result from bi-ionic 110 mm Ca2+-140 mm Cs+ conditions to more physiological Cs+ Tyrode solution (eqn (7)) we obtained *fCa= 0.14. The discrepancy between fCa= 0.01 (eqn (5)) and *fCa= 0.14 is significant. We interpret that *fCa is erroneously high for the following reasons. Firstly, eqns (6) and (7) are based on the assumption of the independence principle of ion diffusion through the channel pore, an assumption that rarely holds true because Cs+ ions can modify the mobility of the Ca2+ ion in the pore (Aidley & Stanfield, 1996). Further, the high [Ca2+]o (110 mm) alters the screening of the surface charges; this effect may have shifted the ICCh-V relationship and Vrev to more positive values (Zholos & Bolton, 1995), and fCa would have been overestimated. Also, the high [Ca2+]o could have unpredictably changed the effective ionic concentration in the mouth of the pore.

Dependence of tonic Δ[Ca2+]c on net inward current

In this study, CCh induced a significant tonic Δ[Ca2+]c only in the presence of a distinguishable ICCh. Depolarization to the reversal potential of ICCh (Vrev= 0 mV, Fig. 3B) abolished Δ[Ca2+]c. The result may be somewhat surprising because the driving force for Ca2+ influx is still high (Vrev - ECa= 120 mV, where ECa is the equilibrium potential for Ca2+). Presumably, the Ca2+ influx at 0 mV was so small that its effect on global [Ca2+]c was fully compensated by the Ca2+ removal processes (eqn (3)). A similar explanation may hold true for the disappearance of Δ[Ca2+]c after substitution of extracellular Cs+ with Na+ (reducing ICCh to ∼5 %, Fig. 1).

The present results indicated that substitution of extracellular cations can modulate the activity (NPo) of single CCh-induced non-selective channels. Substitution of the extracellular 140 mm Na+ by 140 mm Cs+ largely increased NPo and moderately enhanced the unitary conductance (Fig. 2). There is no other report in the literature that demonstrates a similar modulation of the NPo of NSCCCh by monovalent cations. The underlying mechanism has yet to be analysed.

In a first approximation, one may assume that the large tonic ICCh in Cs+ Tyrode solution is proportional to small tonic ICCh as it was recorded in Na+ Tyrode solution, i.e. that the fractional current is 0.01 in both solutions. In Na+ Tyrode solution, tonic ICCh had amplitudes between −10 and −50 pA. The corresponding Ca2+ influx is very small, and presumably it is fully compensated by the concomitant Ca2+ removal processes (eqn (3)). Therefore, one can expect the absence of a tonic Δ[Ca2+]c during the very small tonic ICCh in normal Na+ Tyrode solution. This extrapolation above depends on the assumption that the fractional Ca2+ current of ICCh does not change upon substitution of the charge-carrying cations, an assumption that has not yet been proven.

Physiological importance of CCh-induced Ca2+ influx

In the literature, the existence of CCh-induced Ca2+ influx has been demonstrated only for 110 mm Ca2+ in the bath (Bolton & Kitamura, 1983; Inoue & Isenberg, 1990a; Takai et al. 1997). In guinea-pig ileal SMC, muscarinic stimulation was suggested to stimulate Ca2+ influx through the nifedipine-resistant pathway with the consequence of Ca2+ loading of the SR necessary for the observed Ca2+ oscillations (Komori et al. 1993, 1996). In canine colonic SMC, nicardipine-resistant, ACh-activated conductance was thought to be responsible for a tonic Δ[Ca2+]c (Sato et al. 1994). Despite differences in tissue and species, we cannot offer final and direct evidence supporting the physiological role of Ca2+ influx through NSCCCh. In the gastric myocytes superfused with Tyrode solution containing 150 mm NaCl plus 2 mm CaCl2 (Fig. 1), we measured a small tonic component of CCh-induced inward current at −60 mV that was insensitive to the Ca2+ channel blocker verapamil. Our measurements indicate that this Ca2+ influx is too small to induce a tonic increase in global Δ[Ca2+]c. We suppose that a small Ca2+ influx might increase [Ca2+] in the narrow space between plasmalemma and SR, which could not be detected by the mean fluorescence signal of the whole cytosol (for this dissociation, see Ganitkevich & Isenberg, 1996). Because of its long lasting ‘tonic’ time course, the small tonic CCh-induced Ca2+ influx may contribute to the cellular Ca2+ balance. In experimental conditions with Cs+ Tyrode bath solution, the CCh-induced tonic Δ[Ca2+]c mediates a slow ‘desensitization’ of ICCh via a Ca2+-dependent protein kinase C subtype (Kim et al. 1998). However, the same negative feedback modulation was not confirmed experimentally because of the small amplitude of ICCh in Na+ Tyrode solution.

In summary, our results from gastric myocytes superfused with 140 mm Cs+ Tyrode solution suggest that Ca2+ influx carries ∼1 % of ICCh at physiological [Ca2+]o (2 mm) and at a membrane potential of −60 mV. This is the first quantitative measurement of Ca2+ influx via a G protein-related non-selective cation channel of a smooth muscle cell. The amplitude and significance of CCh-mediated Ca2+ influx in physiological Na+-containing solutions have yet to be studied.

Acknowledgements

This research was supported by a grant from the ‘97 Korea-Germany cooperative science program by KOSEF (No. 975–0700-001–2) and DFG (No. Is/24.18–1).

Ancillary