Carbachol-induced oscillations in membrane potential and [Ca2+]i in guinea-pig ileal smooth muscle cells

Authors


Corresponding author S. Komori: Laboratory of Pharmacology, Department of Veterinary Medicine, Faculty of Agriculture, Gifu University, Yanagido 1-1, Gifu 501-11, Japan. Email: KOMORI_at_CC@DEC.AEDU.GIFU-U.AC.JP

Abstract

  • 1Cytosolic free Ca2+ concentration ([Ca2+]i) and membrane potential were simultaneously recorded from single smooth muscle cells of guinea-pig ileum, using a combination of nystatin-perforated patch clamp and fura-2 fluorimetry techniques.
  • 2Carbachol (CCh, 2 μM) produced oscillatory changes in [Ca2+]i and membrane potential which coincided well in time with each other, and peaks of membrane potential oscillations reached a saturated level of around −7 mV. Thapsigargin (1 μM) abolished these effects of 2 μM CCh. La3+ (3 μM) immediately prevented the discharge of spike potentials, but allowed both on-going oscillatory responses to persist for a while.
  • 3CCh (0.25-0.75 μM) caused membrane potential and [Ca2+]i to oscillate in some 20 % of cells studied. Every membrane potential oscillation was preceded by the discharge of single or multiple spike potentials. The effects of CCh were readily abolished by La3+ (3 μM).
  • 4In cells exhibiting no oscillatory response to 0.25-0.75 μM CCh, an electrically evoked action potential usually generated changes in [Ca2+]i and membrane potential similar to those following spontaneously evoked action potentials, and sometimes it did so only after [Ca2+]i or InsP3 had been slightly elevated by repeatedly evoking action potentials or by increasing CCh concentration in the bath medium.
  • 5The results suggest that in ileal smooth muscle cells, the oscillations of [Ca2+]i and membrane potential arising from muscarinic stimulation result from release of Ca2+ from internal stores and that there is a Ca2+-induced potentiation of coincidently elicited cation channel openings. Under weak muscarinic stimulation, Ca2+ entry upon action potential discharge can trigger such a release of stored Ca2+, resulting in synchronous generation of a large rise in [Ca2+]i and a slow, large membrane depolarization.

In intestinal smooth muscles, muscarinic receptor activation increases [Ca2+]i, and the increased [Ca2+]i activates contractile proteins and regulates the activity of other functional elements such as enzymes and ionic channels. The increase in [Ca2+]i is brought about by the operation of two different mechanisms coupled to muscarinic receptors. One is activation of non-selective cation channels in the plasma membrane, which results in membrane depolarization and initiation or acceleration of action potential discharge (Bolton, 1979b). These membrane potential events stimulate Ca2+ influx through voltage-gated Ca2+ channels (VGCCs) of the L-type, increasing [Ca2+]i (Bolton, 1979b; Tomita, 1981). The other is activation of phospholipase C (PLC), resulting in an increase in the cytosolic InsP3 level (Prestwich & Bolton, 1995). InsP3 activates its receptor, which serves as a Ca2+ release channel, in intracellular Ca2+ stores, and stored Ca2+ is discharged into the cytosol (Kobayashi et al. 1989; Komori & Bolton, 1990, 1991a).

Cytosolic Ca2+ would influence the muscarinic receptor-mediated Ca2+ mobilization through a variety of feedback controls. The opening of muscarinic receptor-operated cation channels is potentiated by a rise in [Ca2+]i due to Ca2+ release from intracellular stores or Ca2+ entry across the plasma membrane (Inoue & Isenberg, 1990b; Pacaud & Bolton, 1991). Such [Ca2+]i rises can also open Ca2+-activated K+ (BKCa) channels with a large conductance (Bolton & Lim, 1989; Komori et al. 1992; Zholos et al. 1994), so that membrane potential is shifted towards hyperpolarization. The activation of VGCCs is inhibited directly by increased [Ca2+]i, through the binding of Ca2+ to the inner mouth of the Ca2+ channel, and indirectly via an unidentified Ca2+-sensitive mechanism (Komori & Bolton, 1991b; Unno et al. 1995). InsP3-gated Ca2+ release channels are under a dual regulation of cytosolic Ca2+; their opening is accelerated as [Ca2+]i increases to a certain level, but inhibited when [Ca2+]i rises higher than this level (Iino, 1990; Bezprozvanny et al. 1991; Finch et al. 1991).

Our recent studies in single smooth muscle cells of guinea-pig ileum revealed that carbachol (CCh) (at muscarinic receptors) is capable of inducing two distinct types of [Ca2+]i oscillation; one is derived from Ca2+ influx associated with action potential discharges, and the other is derived from InsP3-induced release of Ca2+ from intracellular stores (Kohda et al. 1996). Furthermore, voltage clamp studies have shown that CCh produces oscillatory changes in non-selective cation current and BKCa current that are associated with the latter type of [Ca2+]i oscillation (Komori et al. 1993; Zholos et al. 1994). However, little is known as to the change that membrane potential undergoes during such activation of muscarinic receptors and how changes in membrane potential and [Ca2+]i interact with each other.

In the present study, we monitored changes in membrane potential simultaneously with changes in [Ca2+]i during muscarinic receptor activation with CCh in single ileal smooth muscle cells in an attempt to characterize the relationship between the two variables.

The results show that a membrane depolarization occurs synchronously with every [Ca2+]i oscillation induced by CCh, and suggest that individual membrane depolarizations result from repetition of InsP3-induced release of stored Ca2+, which serves to accelerate coincidently elicited cation channel opening. Another major finding is that even when muscarinic receptor activation is insufficient to produce the oscillatory activity, spontaneously generated and electrically evoked action potentials can trigger release of Ca2+ from intracellular stores, resulting in synchronous generation of a rise in [Ca2+]i and membrane depolarization.

METHODS

Preparation of cells

Male guinea-pigs, weighing 350–450 g, were stunned and killed by exsanguination. A 15 cm length of the ileum was removed and divided into five segments (each about 3 cm long). The longitudinal muscle layer of the intestinal segments was peeled from the underlying circular muscle and washed in physiological salt solution (PSS; composition given below). Single isolated smooth muscle cells were prepared from the longitudinal muscle layers using a combination of collagenase (0.2-0.6 mg ml−1) and papain (0.3-0.6 mg ml−1), as described previously (Kohda et al. 1996).

Fura-2 loading of cells

Cells were suspended in a low-Ca2+ (0.5 mM)-containing PSS to which fura-2 acetoxymethyl ester (fura-2 AM; 2 μM) was added, and placed in a dark room at 25°C for 30 min. After the procedure for fura-2 loading, the cell suspension was centrifuged at 700 r.p.m. for 2 min, and the cells were resuspended in the low-Ca2+ PSS without fura-2 AM, placed on coverslips (20 mm in diameter) in a small aliquot and stored in a refrigerator (4°C) in a moist atmosphere. The cells were used for experiments within 8 h after fura-2 loading.

Simultaneous measurements of [Ca2+]i and membrane potential

A shallow chamber (0.5 ml in volume), the base of which was formed by one of the prepared coverslips of fura-2-loaded cells, was mounted on the stage of an inverted fluorescence microscope (Olympus IMT-2, Tokyo, Japan), as described previously (Kohda et al. 1996). The chamber was perfused with 5–10 ml PSS to wash away contaminants in the cell suspension, and then filled with fresh PSS.

Fura-2 fluorescence was measured at room temperature (22–25°C) with two alternating excitation wavelengths (340 and 380 nm) at 100 Hz using an Olympus Ca2+ microspectrometric system with a × 40 objective (OSP-3 model). Fluorescent light was collected from the whole area of a single cell of interest and counted by a photomultiplier tube through a bandpass filter (510 ± 30 nm). Counts of the fluorescence at 340 nm (F340) and 380 nm excitation (F380) were sampled at 20 Hz. [Ca2+]i was calculated from the F340/F380 ratio according to the formula (Grynkiewicz et al. 1985):

display math

The maximum fluorescence ratio (Rmax) was obtained by exposing cells to a 50 μM concentration of a Ca2+ ionophore, ionomycin, in PSS (containing 2 mM Ca2+). Immediately after determining Rmax, the solution was replaced with a Ca2+-free solution containing 50 μM ionomycin and 5 mM EGTA, and the minimum fluorescence ratio (Rmin) was determined. The fluorescence ratio of F380 in Ca2+-free solution to that in PSS (β) was determined in the experiments mentioned above. Mean values for Rmin, Rmax and β were 0.44 ± 0.01, 5.64 ± 0.19 and 4.07 ± 0.23 (n= 7), respectively. These mean values and the dissociation constant for the Ca2+-fura-2 complex (Kd), taken to be 224 nM (Grynkiewicz et al. 1985), were used for calculation of [Ca2+]i. Ratios were stored on a disk and a PCM data recorder (RD-111T, TEAC, Musashino City, Tokyo, Japan).

During fluorescence measurements, fura-2-loaded cells were also held under whole-cell patch clamp using the nystatin-perforated patch technique, as described previously (Kohda et al. 1996). This technique has the advantage that it prevents run-down of the activity of ionic channels such as voltage-gated Ca2+ channels and diffusion of intracellular functional proteins into the patch pipette (Wakamori et al. 1993; Fleischmann et al. 1996). Patch pipettes filled with a KCl-based solution (composition given below), with a resistance of 4–6 MΩ were used for recording membrane potential using a patch clamp amplifier (CEZ-2400, Nihon Kohden, Tokyo, Japan). Measurements of membrane potential were stored on a PCM data recorder (RD-111T, TEAC) and replayed onto a thermal array recorder (RTA-1100M, Nihon Kohden) for analysis and illustration.

Values in the text are given as means ±s.e.m. with the number of cells (n) used for measurements. Statistical significance was tested using Student's unpaired t test and differences were considered significant when P < 0.05.

Solutions and drugs

The PSS used in the experiments had the following composition (mM): NaCl, 134; KCl, 6; CaCl2, 2; MgCl2, 1.2; glucose, 14 and Hepes, 10.5 (titrated to pH 7.2 with NaOH). The KCl-based pipette solution had the following composition (mM): KCl, 134; Hepes, 10.5 (titrated to pH 7.2 with KOH), to which nystatin dissolved in DMSO (4 mg 0.1 ml−1) was added to give a final concentration of 0.2 mg ml−1.

Drugs used were thapsigargin, nystatin and nicardipine, all of which were purchased from Sigma; carbachol, from Wako (Osaka, Japan); and fura-2 AM, from Dojin Kagaku (Kumamoto, Japan). All other reagents were of the highest grade commercially available.

Application of drugs was made by replacing the solution bathing the cells with drug-containing solution 5–7 times within 10 s.

RESULTS

CCh-induced oscillations in membrane potential and [Ca2+]i

Cells exhibited no electrical activity with a resting membrane potential of −50.6 ± 0.8 mV and a steady [Ca2+]i of 42.0 ± 1.3 nM (n= 52) in the normal PSS.

Figure 1 shows simultaneous recordings of changes in membrane potential and [Ca2+]i in response to 2 μM CCh. The [Ca2+]i change consisted of an oscillatory component and a sustained component. The sustained component developed slowly and reached a level 20–70 nM above the initial steady level. Individual [Ca2+]i oscillations lasted some 5 s, superimposed on the sustained component, and the interoscillation interval was roughly constant for each particular cell. The mean oscillation frequency in different cells was 0.13 ± 0.01 Hz (n= 33). The amplitude of the oscillations, measured as the difference between the basal and peak [Ca2+]i, was greatest for the first oscillation (372.3 ± 29.0 nM, n= 33) and then decreased with time towards a stable level. The rate of decline and the stable level attained varied from one cell to another. These characteristics of the oscillatory [Ca2+]i response were the same as those of the [Ca2+]i oscillations arising from InsP3-induced release of Ca2+ from intracellular stores as previously reported (Komori et al. 1993; Kohda et al. 1996).

Figure 1.

Oscillations in membrane potential and [Ca2+]i induced by 2 μM CCh in single smooth muscle cells of guinea-pig ileum

Membrane potential (upper traces) and [Ca2+]i (lower traces) were simultaneously measured using a combination of nystatin-perforated whole-cell patch clamp and fura-2 fluorimetry techniques. A and B, records from two different cells. In A, a pair of records indicated by a and b are also shown on a time-extended scale. Note that the peak of membrane potential oscillations does not exceed the zero potential.

Oscillatory changes in membrane potential with a sustained component were also elicited, and coincided well in time with the oscillatory changes in [Ca2+]i (see Fig. 1). The sustained depolarization developed slowly and attained a plateau level during the CCh application, the amplitude of which varied from several millivolts up to 30 mV in different cells. Some oscillations of membrane potential, especially those generated in the early part of the duration of CCh application, carried a single spike potential or multiple spike potentials on their depolarizing phase. As illustrated in Fig. 1A, a single spike potential was discharged with an overshoot potential (up to 25 mV). Its repolarization usually ended at a level close to the peak potential of the subsequently elicited slow depolarization (hereafter referred to as the oscillation peak). In any oscillatory response of membrane potential, the oscillation peak increased gradually with repetition of oscillation in some early cycles and then became steady or declined slowly with time. The mean peak potential was −16.3 ± 1.0 mV (n= 33) in the first oscillation and somewhat negative to 0 mV (−7.0 ± 0.5 mV, n= 33) in the subsequent well-developed oscillations. The saturated peak potential was close to the reversal potential of the CCh-evoked non-selective cation channel current in the same cell type (−9 mV, Bolton, 1972; −7.4 mV, Bolton et al. 1981).

The oscillations of membrane potential coincided well with those of [Ca2+]i in their generation time, but their mutual relationship in peak amplitude was less consistent. This was especially apparent between the first two oscillations in membrane potential and [Ca2+]i: the first oscillation had a higher peak for [Ca2+]i but a lower, or less positive, peak for depolarization, compared with the subsequent oscillation (Fig. 1). In some cells, the peak of the [Ca2+]i oscillations continued to decline, while the peak of the corresponding depolarizations was maintained at a steady level.

Effect of thapsigargin on the CCh-induced oscillations in membrane potential and [Ca2+]i

Treatment with thapsigargin, an inhibitor of Ca2+-ATPase in intracellular Ca2+ stores, at 1 μM for 1.5-2 min to deplete CCh-releasable Ca2+ stores (Komori et al. 1996), rendered cells unable to respond to 2 μM CCh with the oscillatory activity. In three out of six thapsigargin-treated cells, only small irregular fluctuations in membrane potential occurred, with a sustained increase in [Ca2+]i of the order of a few tens of nanomolar (Fig. 2A). In the remaining three cells, a sustained membrane depolarization by 16–30 mV superimposed with small irregular fluctuations was elicited, as shown in Fig. 2B. On the other hand, [Ca2+]i was increased in a virtually biphasic manner, with an initial transient increase by 100–160 nM followed by a sustained increase with a smaller amplitude. The membrane depolarization gradually declined along with the attenuation of the sustained [Ca2+]i increase even in the continued presence of CCh.

Figure 2.

Changes in membrane potential and [Ca2+]i produced by 2 μM CCh after treatment with thapsigargin

Thapsigargin (1 μM) was applied 1.5 min before CCh application. A and B, records of CCh-induced changes in membrane potential (upper traces) and [Ca2+]i (lower traces) from two different cells. Note failure of CCh to produce any oscillatory response. See text for details.

Thapsigargin alone had little effect on the resting membrane potential, and caused an increase in [Ca2+]i by up to 30 nM (n= 6) which attained a peak within 30–60 s and then gradually declined towards the initial resting level.

These results imply that after functional removal of Ca2+ stores, 2 μM CCh is incapable of generating the oscillatory changes in membrane potential and [Ca2+]i.

Effect of La3+ on the CCh-induced oscillations in membrane potential and [Ca2+]i

La3+ (3 μM), an inorganic Ca2+ channel blocker, when applied to cells exhibiting oscillatory responses to 2 μM CCh, arrested on-going oscillations in membrane potential and [Ca2+]i (n= 5). As shown in Fig. 3A, the effect of La3+ was exerted after several oscillations in [Ca2+]i with gradually decreased amplitudes and in membrane potential with saturated peak amplitude. In contrast, spike potentials, if there, were not observed after La3+ application.

Figure 3.

Effect of La3+ on CCh-induced oscillations in membrane potential and [Ca2+]i

A, records of CCh (2 μM)-induced changes in membrane potential (upper trace) and [Ca2+]i (lower trace) before and after application of 3 μM La3+. Application of CCh and La3+ is indicated by bars below the [Ca2+]i recording trace. B, records when CCh was used at 0.5 μM. A pair of records indicated by a and b is also shown on a time-extended scale. A and B are from different cells. Note that both on-going oscillations in membrane potential and [Ca2+]i induced by 2 μM CCh persisted for a while after application of La3+, whereas those induced by 0.5 μM CCh stopped immediately. See text for details.

Effect of lower concentrations of CCh on membrane potential and [Ca2+]i

When CCh was applied at concentrations of 0.25-0.75 μM, most of the cells used (14 out of 18) responded with a sustained increase in [Ca2+]i of up to 40 nM and a sustained membrane depolarization of up to 12 mV. However, in the remaining four cells, synchronous oscillations in [Ca2+]i and membrane potential were elicited with varied interoscillation intervals, as shown in Fig. 3B. The membrane potential oscillations had a peak potential of −9.8 ± 2.8 mV (n= 4) and a total duration of 3–5 s (n= 4), and they invariably carried a single spike potential on their depolarizing phase. In these respects, they were indistinguishable from the 2 μM CCh-induced membrane potential oscillations that carried spike potentials. However, the spike potential differed from that in the response to 2 μM CCh in that its repolarization after passing over an overshoot potential attained a substantially more negative level than the oscillation peak (cf. Figs 3B and 1A); namely, there was an appreciable time lag between the peaks of the spike potential and oscillation. In relation to this, the rising phase of [Ca2+]i oscillations was divided into two parts, an initial rapid component and a slower component (see Fig. 3B). These oscillatory activities were abolished immediately after 3 μM La3+ was applied (Fig. 3B).

These results suggest that during exposure to these low concentrations of CCh, the spike potential discharge may be a prerequisite for the generation of such a membrane depolarization and a rise in [Ca2+]i.

Effect of electrically evoked action potentials on membrane potential and [Ca2+]i in the presence of CCh

To test the possibility that the oscillations in membrane potential and [Ca2+]i in the presence of a low concentration of CCh are under the command of the spike activity, the following experiments were carried out.

In cells exhibiting no oscillatory activity during exposure to 0.25-0.75 μM CCh, depolarizing current pulses (20 or 30 pA in intensity and 100 ms in duration) were applied to evoke single action potentials. As shown in Fig. 4A, the evoked action potential was followed by a slow depolarization in synchrony with a large rise in [Ca2+]i (see also Fig. 4B), although it was without effect before exposure to CCh (Fig. 4Aa and Ba). The slow depolarization was comparable in shape and amplitude with the 2 μM CCh-induced oscillation in membrane potential, which lasted 5 s or so and carried one or more spike potentials on the depolarizing phase (Fig. 4B). The mean peak potential of the slow depolarizations was −7.6 ± 1.2 mV (n= 7) with a range from −12.5 to −3.2 mV in different cells.

Figure 4.

Effect of thapsigargin on changes in membrane potential and [Ca2+]i induced by electrically evoked action potentials in the presence of 0.5 μM CCh

A, records of changes in membrane potential (top trace) and [Ca2+]i (bottom trace) induced by the evoked action potentials before and after CCh application and after application of 1 μM thapsigargin in the continued presence of CCh, as indicated by bars below the [Ca2+]i recording trace. Depolarizing current pulses (20 pA in intensity and 100 ms in duration; middle trace) were used to evoke single action potentials. B, time-expanded recordings of traces indicated by a, b and c in A. Note that both membrane depolarization and [Ca2+]i increase subsequent to the evoked action potential in the presence of CCh were selectively and markedly inhibited by thapsigargin.

The increase in [Ca2+]i following the evoked action potential occurred in a biphasic manner; an initial small increase of 47.3 ± 5.0 nM (n= 7) followed by a slow, large increase of 200.0 ± 18.1 nM (n= 7) with small irregular fluctuations. The initial, small increase in [Ca2+]i was closely associated with the evoked action potential, and the subsequent large increase in [Ca2+]i with the later depolarization (Fig. 4B). The mean value of 47.3 ± 5.0 nM for the initial, small [Ca2+]i increase was not significantly different from the action potential-triggered [Ca2+]i rise before exposure to CCh (40.9 ± 3.7 nM, n= 7; see Fig. 4Aa and Ba).

The membrane depolarization and [Ca2+]i increase subsequent to the evoked action potential were both markedly reduced by thapsigargin (1 μM) with little or no change of the evoked action potential (see Fig. 4A and B). The reduced depolarization ranged from 3.6 to 10.9 mV in four different cells. The remaining [Ca2+]i increase was 47.4 ± 6.7 nM (n= 4), which was similar in amplitude to the action potential-triggered [Ca2+]i rise before exposure to CCh, but somewhat longer in duration, probably because of the virtual lack of the ability of intracellular Ca2+ stores to take up Ca2+ in the presence of thapsigargin (Kohda et al. 1997).

As shown in Fig. 5A, La3+ (3 μM) immediately blocked action potential discharge in response to depolarizing current application, leaving a small membrane depolarization due to the passive membrane property (n= 3, see also Fig. 5B). The small depolarization was not followed by any appreciable change in membrane potential and [Ca2+]i. Similar results were also obtained with nicardipine (1 μM, n= 2), an organic Ca2+ channel blocker (data not shown).

Figure 5.

Effect of La3+ on the membrane depolarization and [Ca2+]i rise following electrically evoked action potentials in the presence of 0.5 μM CCh

A, records of changes in membrane potential (top trace) and [Ca2+]i (bottom trace) following the evoked action potentials before and after application of 3 μM La3+ in the continued presence of CCh. B, time-expanded recording traces of the membrane potential changes indicated by a and b in A. Note that La3+ abolished both the large membrane depolarization and [Ca2+]i rise as well as the evoked action potential and its concomitant [Ca2+]i rise. See text for details.

The results strongly suggest that under the conditions in which muscarinic receptors are activated by 0.25-0.75 μM CCh, action potential discharge can trigger release of Ca2+ from intracellular stores, leading to the synchronous generation of a slow, large depolarization and a large rise in [Ca2+]i, and that this effect is attributable to Ca2+ entry through voltage-gated Ca2+ channels.

Possible mechanism underlying the action potential-triggered Ca2+ release from intracellular stores

To see whether the release of stored Ca2+ upon action potential discharge involves a mechanism through which InsP3 and Ca2+ act as co-agonists to open InsP3-gated Ca2+-release channels (Iino, 1990; Bezprozvanny et al. 1991; Finch et al. 1991), the following experiments were carried out.

A cell was exposed to two different concentrations of CCh (0.25 and 0.5 μM) in sequence to further accelerate muscarinic receptor-mediated formation of InsP3 (Prestwich & Bolton, 1995). As shown in Fig. 6A, in the presence of 0.25 μM CCh, an electrically evoked action potential was followed by a small depolarization of 7.3 mV and a rise in [Ca2+]i by 45.2 nM, which was no more than that triggered before CCh application. After CCh concentration was increased to 0.5 μM, the depolarization increased to 39.7 mV and the [Ca2+]i rose by 170.6 nM. Similar results were obtained in five other cells with a combination of two different CCh concentrations ranging from 0.25 to 0.75 μM.

Figure 6.

Effect of increases in CCh concentration and Ca2+ influx on changes in membrane potential and [Ca2+]i following an electrically evoked action potential

Depolarizing current pulses (30 pA in intensity and 100 ms in duration; middle traces) were used to evoke action potentials. A, records of evoked action potential-induced changes in membrane potential (top trace) and [Ca2+]i (bottom trace) before (a) and after application of 0.25 μM (b) and then 0.5 μM CCh (c). B, records of membrane potential and [Ca2+]i changes induced by a single evoked action potential (a) and two evoked action potentials with a 100 ms interval (b) before application of 0.5 μM CCh and of those (c and d, respectively) after its application. A and B are from different cells. Note that the increases of CCh concentration from 0.25 to 0.5 μM (A) and of Ca2+ influx by repeated action potential discharge (B) enhanced both the membrane depolarization and [Ca2+]i increase induced by a single pulse-evoked action potential. See text for details.

In another series of experiments, action potentials were repeatedly evoked to increase Ca2+ entry in the presence of a low concentration of CCh (0.25 or 0.5 μM). As shown in Fig. 6Bd, subsequent to two successive action potentials evoked at an interval as short as 100 ms, a slow depolarization of 35.5 mV and a rise in [Ca2+]i by 132.0 nM were generated. This [Ca2+]i rise was greater than the control (89.5 nM) measured immediately before CCh application (cf. Fig. 6Bb and Bd).

The results suggest that the InsP3-Ca2+ co-agonist mechanism is operated in the action potential-triggered release of Ca2+ from intracellular stores.

The refractory period of the action potential-triggered release of Ca2+ from intracellular stores

A pair of current pulses (P1 and P2) was applied at various intervals (up to 20 s) to cells which were not showing oscillatory responses to 0.5 μM CCh. Action potential discharge in response to application of P1 was effective in generating release of stored Ca2+ as judged by the [Ca2+]i rise and membrane depolarization. Interestingly, there was a period during which an action potential of full size evoked by P2 was followed by a very suppressed depolarization and [Ca2+]i rise compared with those subsequent to the P1-evoked action potential (Fig. 7Aa). As the interval between P1 and P2 was increased further, both the membrane depolarization and [Ca2+]i rise subsequent to the P2-evoked action potential increased until they were generated with full amplitudes (Fig. 7Ab and Ac). The amplitudes of the depolarization and [Ca2+]i rise expressed as percentages of those subsequent to the P1-evoked action potential were plotted against the interpulse interval. The plots obtained from four different experiments are shown with different symbols in Fig. 7B and C for the [Ca2+]i rise and the depolarization, respectively. The data plots show that full recovery of the action potential-triggered release of stored Ca2+ occured in 5–10 s.

Figure 7.

Effects of paired current pulses on membrane potential and [Ca2+]i in the presence of CCh

A pair of current pulses (30 pA in intensity and 100 ms in duration) was applied to evoke two successive action potentials at varied intervals in the presence of 0.5 μM CCh. A, records of changes in membrane potential (top traces) and [Ca2+]i (bottom traces) in response to a pair of current pulses (middle traces) at interpulse intervals of 3.2 (a), 6 (b) and 20 s (c). Note that as the interpulse interval increased, membrane depolarization and [Ca2+]i rise induced by the second evoked action potential increased. B and C, plots of amplitudes of the [Ca2+]i rise (B) and membrane depolarization (C) induced by the second action potential against the interpulse interval. The amplitudes of both variables (ordinates) are expressed as percentages of those induced by the first evoked action potential (P1). Data points with different symbols are derived from four different cells.

DISCUSSION

The present results show that activation of muscarinic receptors with CCh (2 μM) causes [Ca2+]i and membrane potential to oscillate synchronously with each other in single smooth muscle cells of guinea-pig ileum. The [Ca2+]i oscillation has already been ascribed to repetition of InsP3-induced release of Ca2+ from intracellular stores (Komori et al. 1993; Kohda et al. 1996). How is the membrane potential oscillation elicited? Muscarinic receptor activation can also induce opening of non-selective cation (NSC) channels through a pertussis toxin-sensitive G-protein, leading to membrane depolarization (Inoue & Isenberg, 1990a; Komori et al. 1992). The muscarinic effect has been demonstrated to be strongly potentiated by a rise in [Ca2+]i through a mechanism by which the G-protein-NSC channel link is rendered Ca2+ sensitive (Inoue & Isenberg, 1990a, b; Pacaud & Bolton, 1991). If this is so, the cation channels will be activated to varying extents as [Ca2+]i changes. In other words, the membrane potential can be expected to change along with changes in [Ca2+]i. In fact, membrane potential was [Ca2+]i dependent and the membrane depolarizations had a fixed peak potential close to the reversal potential described for the cation channel current (−9 mV, Bolton, 1972; −7.4 mV, Bolton et al. 1981). Therefore, the membrane depolarization is the result of [Ca2+]i-potentiated activation of NSC channels.

Ca2+-activated K+ (BKCa) channels in the cell should also be opened by the increased [Ca2+]i arising from muscarinic receptor activation (Komori et al. 1992; Zholos et al. 1994), and their opening would act to reduce the depolarizing effect of the NSC channels. The present observation of the [Ca2+]i-dependent, depolarizing change in membrane potential implies that the contribution of BKCa channel activation to the determination of the overall effect on the membrane potential is minor.

The relationship between the [Ca2+]i rise and the membrane depolarization in the first oscillation cycle would result from a small activation of NSC channels because of a slow increase in Ca2+ sensitivity of the G-protein-NSC channel link after muscarinic receptor stimulation (Komori et al. 1993). After the Ca2+ sensitivity of this link had been sufficiently increased, the NSC channel, like the BKCa channel, would be activated in an [Ca2+]i-dependent manner. Thus, the incompatibility of the peak amplitude between the [Ca2+]i rise and the membrane depolarization in some cycles immediately before termination of their oscillations (see Figs 1 and 3A) can be readily explained by the idea that NSC channel activation contributes predominantly to determination of the change in membrane potential (see above). However, the idea might prove incorrect if NSC channel activation is extraordinarily different in [Ca2+]i dependency from BKCa channel activation. The NSC channel has been shown to be activated over a relatively wide range of [Ca2+]i, whereas the BKCa channel seems to be activated in a narrow range of [Ca2+]i (Pacaud & Bolton, 1991; Komori et al. 1993; Zholos et al. 1994).

Ca2+ entry may occur through activated NSC channels and/or through voltage-gated Ca2+ channels if action potentials are discharged. Even if this was so, the Ca2+ entry should be much smaller in amount than Ca2+ released from intracellular stores. This may also help to explain why the [Ca2+]i oscillations with gradually decreased amplitudes occurred even when the membrane potential oscillations were elicited with a constant peak potential, whether or not they carried spike potentials. This idea is supported by the fact that the cation channels underlying the membrane potential oscillations are not measurably permeable to Ca2+ (Pacaud & Bolton, 1991).

It is of particular interest that during muscarinic receptor activation with lower concentrations (0.25-0.75 μM) of CCh, spontaneously generated and electrically evoked action potentials were followed by a rise in [Ca2+]i and a membrane depolarization comparable in shape and size with 2 μM CCh-induced oscillations in [Ca2+]i and membrane potential. Neither the [Ca2+]i rise nor the membrane depolarization were induced after prevention of action potential discharge by La3+ or nicardipine, and both were abolished after functional removal of intracellular Ca2+ stores by thapsigargin. It is, therefore, highly probable that these responses result from Ca2+ release from intracellular Ca2+ stores triggered by Ca2+ entry upon action potential discharge.

The release of Ca2+ from intracellular stores triggered by Ca2+ entry resulting from the activation of L-type Ca2+ channels might be a manifestation of the Ca2+-induced Ca2+ release (CICR) mechanism: Ca2+ entry activates ryanodine-sensitive Ca2+-gated Ca2+ release channels to discharge stored Ca2+. However, our previous study in the same cell type (Kohda et al. 1997) shows that the CICR mechanism does not operate in response to Ca2+ entry upon action potential discharge.

In some cells, the action potential evoked by depolarizing current application failed to trigger Ca2+ release from intracellular stores giving rise to such a [Ca2+]i rise and membrane depolarization in the presence of CCh. This failure seems likely to be due to an insufficient accumulation of InsP3, since the action potential became effective after a small elevation of CCh concentration in the bathing medium. Similar results were obtained after repetition of action potential discharge, which caused an increase in Ca2+ entry (see Fig. 6). This can be well explained by the idea that InsP3 and Ca2+ act as co-agonists to switch on activation of InsP3-gated Ca2+ release channels, and a massive release of stored Ca2+ occurs (Iino, 1990; Bezprozvanny et al. 1991; Finch et al. 1991).

Once the action potential-triggered release of stored Ca2+ was elicited, the next release was suppressed, and full recovery occurred within 10 s. Thus, the underlying mechanism of this Ca2+ release from intracellular stores is thought to have a refractory period of several seconds. This is not inconsistent with the properties of InsP3-gated Ca2+ release channels reported previously; the channels are inactivated when [Ca2+]i rises above a certain level (Iino, 1990; Bezprozvanny et al. 1991; Finch et al. 1991), and the response, reflecting release of stored Ca2+, to flash-liberated InsP3 from caged InsP3 is severely inhibited until full recovery occurs, even after repletion of Ca2+ stores (Zholos et al. 1994).

Spontaneous discharge of action potentials is usually recorded from tissue preparations using a microelectrode (e.g. Bolton, 1972). However, most of the single cells used exhibited no spontaneous electrical activity. The difference may be due to the use of different preparations and temperatures (22–25°C for the present study and 35°C for that of Bolton). The action potential-triggered release of Ca2+ from intracellular stores found in the present study is considered to play an important role in the mechanical activity of intestinal smooth muscle. Constituent smooth muscle cells receive cholinergic innervation from the enteric plexuses, and they are exposed to acetylcholine released, whether spontaneously or in response to nerve impulses (Paton et al. 1971; Takewaki et al. 1975). Under these conditions, action potentials in the smooth muscle cells would cause operation of this type of stored-Ca2+ release mechanism by which the subsequent tension development is increased. The mechanism may act as an amplifier in the excitation-contraction coupling. This would account for some of the inhibitory effect of atropine on the spontaneous mechanical activity in intestinal smooth muscle (Bolton, 1979a). If the slow, large membrane depolarization associated with the release of Ca2+ from intracellular stores occurred in concert in cells of some regions of this tissue, it could spread electrotonically and play a role in the periodic change in the electrical activity of the tissue.

In smooth muscles including guinea-pig ileum, M2 and M3 subtypes of muscarinic receptors have been suggested to couple to NSC channel opening and to Ca2+ release from intracellular stores via InsP3 formation, respectively, (Candell et al. 1990; Zhang & Buxton, 1991; Cuq et al. 1994; Wang et al. 1997; Zholos & Bolton, 1997; Komori et al. 1998). If so, the present study provides evidence that there is cross-talk between signal pathways of M2 and M3 subtypes. Activation of the M2-NSC channel system causes membrane depolarization and action potential discharge to convey Ca2+ to the cytosol; the Ca2+ in turn potentiates M3 receptor-mediated activation of InsP3-gated Ca2+ release channels, so eliciting a massive release of stored Ca2+ into the cytosol. The increased [Ca2+]i further accelerates the cation (NSC) channel opening to produce a larger membrane depolarization and further Ca2+ influx. With an appropriately strong muscarinic receptor stimulation, the M3 receptor-mediated release of stored Ca2+ occurs repeatedly without Ca2+ as a trigger and the M2-NSC channel system is modified in proportion to the [Ca2+]i change. The contractile response of smooth muscle to muscarinic agonists is considered to be mediated by M3 receptors but not M2 receptors, which account for the major population of muscarinic receptors (Caulfield, 1993; Eglen et al. 1996). However, M2 receptors seem likely to play a more important role in mediating the contractile response than has been thought so far.

In summary, muscarinic receptor stimulation produces synchronous oscillations of [Ca2+]i and membrane potential in intestinal smooth muscle cells. Both oscillatory responses involve InsP3-induced release of Ca2+ from intracellular stores as a primary mechanism. Operation of the stored-Ca2+ release mechanism is under the command of action potential discharge or takes place by itself, depending on the concentration of agonist and the fractional occupancy of muscarinic receptors by the agonist. The oscillations of membrane potential are brought about by potentiation by Ca2+ of activation of muscarinic receptor-operated cation channels.

Ancillary