The α₁-subunit of smooth muscle Ca²⁺ channel preserves multiple open states induced by depolarization

Abstract

1. The cloned α₁-subunits of the smooth muscle Ca²⁺ channel (α_1C-b) from rabbit lung were expressed in Chinese hamster ovary cells. The effect of large depolarizations was examined using cell-attached patch clamp techniques.

2. After large, long-duration depolarizations (to +80 mV, 4 s), the cloned smooth muscle Ca²⁺ channels were still open, and also showed slow channel closure upon repolarization. The sum of unitary channel currents revealed that the tail current seen after large conditioning depolarizations had a slower deactivation time constant compared to that seen when the cell membrane was depolarized briefly with a test step (to +40 mV), suggesting that large depolarizations transform the conformation of the Ca²⁺ channels to a second open state.

3. The decay time course of the tail current induced by large conditioning depolarizations was prolonged by reducing the negativity of the repolarization step, and vice versa.

4. Using the slow deactivating characteristic, the current-voltage relationship was directly measured by applying a ramp pulse after a large depolarization. Its slope conductance was approximately 26 pS.

5. Since the patch pipettes contained Ca²⁺ agonists, the transition of the Ca²⁺ channel conformation to the second, long open state during a large depolarization was distinct from that caused by Ca²⁺ agonists, suggesting that the cloned α₁-subunits of smooth muscle Ca²⁺ channels preserve the characteristic features seen in native smooth muscle Ca²⁺ channels.

6. In addition, when skeletal muscle β-subunits were coexpressed with the α₁-subunits, the long channel openings after large, long-duration depolarizations were frequently suppressed. This phenomenon could be explained if the skeletal muscle β-subunits increased the inactivation rate during the preconditioning depolarization.

In smooth muscle, it is well known that spike activities are largely due to activation of voltage-sensitive Ca²⁺ channels (Brading et al. 1981). Sustained Ca²⁺ influx seen during long-term depolarization of the cell membrane is also considered to be through these voltage-sensitive Ca²⁺ channels (Bolton, 1979). Since both phenomena are significantly suppressed by Ca²⁺ antagonists, (dihydropyridine-sensitive) L-type Ca²⁺ channels seem to cover a wide range of smooth muscle behaviour: from transient to persistent events.

Previously, we have demonstrated the presence of multiple open states in smooth muscle L-type Ca²⁺ channels, in guinea-pig urinary bladder, taenia caeci and gastric antrum (Nakayama et al. 1996). During a large, long-duration depolarization, the conformation of Ca²⁺ channels is transformed to a second open state in which these channels deactivate very slowly upon repolarization (Nakayama & Brading, 1993a). On the other hand, it is well known that dihydropyridine Ca²⁺ channel agonists, such as Bay K 8644, prolong Ca²⁺ channel opening during depolarization (Hess et al. 1984). In our previous experiments, large depolarizations further slowed the deactivation time course of smooth muscle Ca²⁺ channels, even in the presence of Ca²⁺ channel agonists. It was thus suggested that the second open state seen in smooth muscle differs from the dihydropyridine Ca²⁺ channel agonist-induced gating mechanism (Nakayama & Brading, 1995a, 1996). Moreover, when the large depolarizing step was applied for up to 5 s, the Ca²⁺ channels in the second open state did not inactivate (Nakayama & Brading, 1993b, 1995b). Generally, smooth muscles contract slowly, and tonic contraction is one of their characteristic features. It is interesting to note that the transition of Ca²⁺ channels into the second, long open state occurs in smooth muscles with this characteristic feature.

Molecular biological studies have revealed that Ca²⁺ channels consist of several subunits, and also that Ca²⁺ channels are diverse in numerous cells and tissues. Thus, we now know of several kinds of genetically different α₁-subunits, which form the essential parts of voltage-dependent channels: the channel pore, the ionic filter and voltage sensors. α₁-Subunits of L-type Ca²⁺ channels derived from cardiac and smooth muscles are classified as Ca (α_1C-a) and Cb (α_1C-b), respectively. These channel proteins are splice products of the same gene, which is distinct from the gene for skeletal muscle α₁-subunits (summarized by Hofmann & Klugbauer, 1996).

In the present study, we examined whether the smooth muscle α₁-subunit of L-type Ca²⁺ channels preserves the characteristics of the channel behaviour seen in our previous experiments (long channel opening; non-inactivation during depolarization), which could thus be considered a common feature of smooth muscle Ca²⁺ channels. It is likely that large conditioning depolarizations activate ionic conductances other than voltage-sensitive Ca²⁺ channels in whole-cell recordings of tail currents. To rule out this possibility we recorded unitary Ca²⁺ channel currents. Cell-attached recordings of Chinese hamster ovary (CHO) cells expressing the α₁-subunits showed that the closure kinetics upon repolarization is significantly slowed by preconditioning large depolarizations, even in the presence of a high concentration of Bay K 8644. This result reinforces our deduction from previous experiments performed in smooth muscle cells that the long opening of Ca²⁺ channels seen after large depolarizations is due not to the presence of multiple types of Ca²⁺ channels, but to conversion of the channel conformation of a single type of Ca²⁺ channel to a second open state (Nakayama & Brading, 1993a) which is distinct from the channel gating induced by Ca²⁺ channel agonists (Nakayama & Brading, 1995a).

METHODS

Cell preparation

The details of the methods used for stable expression of cloned smooth muscle α₁-subunits (α_1C-b) and skeletal β-subunits (β_1a) of dihydropyridine-sensitive (L-type) Ca²⁺ channels in CHO cells, and for culturing these cells have been described previously (Welling et al. 1993). The expression vectors for α₁- (p91α12b) and β-subunits (pKNHB1) were transfected into CHO cells by electroporation. CHO cells were cultured in Dulbecco's modified Eagle's medium containing 10 % dialysed fetal bovine serum, non-essential amino acids, streptomycin (30 μg ml⁻¹) and penicillin (30 units ml⁻¹). CHO cells expressing both α₁- and β-subunits were selected and maintained via their resistance to neomycin.

Electrical recording

The methods used for electrophysiological experiments were essentially the same as previously described (Nakayama & Brading, 1996). Unitary Ca²⁺ channel currents were recorded in the cell-attached mode using a patch clamp amplifier (Axopatch 200A, Axon Instruments). The experiments were performed at room temperature (22-26°C). A cut-off frequency of ∼1.2 kHz (8 pole Butterworth filter) was applied. The dwell time of the sampling was set to be 50-100 μs. When a Ba²⁺-rich solution was used to fill the patch pipette, the pipette resistance was in the region of 5 MΩ. When single channel currents were measured, the seal resistance between the pipette and cell membrane was 10-20 GΩ. Before establishment of the seal, the reference potential was set to be the zero current potential of the patch pipette (with a high-K⁺ solution in the bath). Capacitive surge was partially compensated electrically. Unless otherwise stated, the voltage of the patch membrane was clamped at -80 mV.

Test voltage steps to +40 mV were applied. In our previous study using the whole-cell patch clamp recording mode, smooth muscle Ca²⁺ channels were almost maximally activated at +20 mV (Nakayama & Brading, 1995b). In this study, since superfusion of the cells with high-K⁺ solutions led to some uncertainty about the membrane potential, a larger potential (+40 mV) was used to achieve maximal activation of Ca²⁺ channels.

The numerical data were expressed as means ±s.d.

Drugs and solutions

The pipette solution had the following composition (mM): BaCl₂ 100; Hepes/Tris 11.8 (pH 7.4 at 24°C). The pipette solution also contained 2 μM Bay K 8644. The composition of ‘Ca²⁺-free, high-K⁺’ bathing solution was as follows (mM): NaCl 30.9; KCl 100; MgCl₂ 3.6; glucose 11.8; Hepes/Tris 11.8 (pH 7.4-7.5 at 24°C). The extracellular Na⁺ was not completely removed, because intracellular pH and other ionic environments may be significantly altered under Na⁺-free conditions (e.g. Nakayama et al. 1994) and such ionic environments may consequently affect the kinetics of the Ca²⁺ channels (e.g. effects of pH: Kaibara & Kameyama, 1988; Iino et al. 1994; effects of Mg²⁺: Hartzell & White, 1989; McHugh & Beech, 1996).

(±)Bay K 8644 was purchased from Calbiochem. Its stock solution (1 mM, dissolved with ethanol) was kept cool and protected from light. The drug was diluted just before use.

RESULTS

Prolongation of channel opening after large depolarizations in cloned smooth muscle α₁-subunits

The effect of large conditioning depolarizations was examined in cloned smooth muscle α₁-subunits of voltage-sensitive Ca²⁺ channels (α_1C-b) stably expressed in CHO cells. Unitary Ca²⁺ channel currents were recorded in the cell-attached mode of the patch clamp technique. Ba²⁺ (100 mM) was used as a charge carrier. The patch membrane was normally kept at -80 mV (holding potential). A test potential of +40 mV (7.5 ms or 15 ms) was applied at 30 s intervals, and every other test potential was preceded by a large conditioning depolarization (+80 mV, 4 s) in order to rule out systematic errors. The test potential (+40 mV) was chosen to induce a maximal degree of activation. The deactivating characterstics were assessed by applying a repolarization step to -60 mV preceded by the test step. To increase Ca²⁺ channel availability, Bay K 8644 was added to the patch pipette solution: channel opening was rare when it was not present.

Figure 1A shows examples of current traces obtained from recordings from the same membrane patch. The channel closure was significantly slowed after the large conditioning depolarization. In this membrane patch, the mean closing times after repolarization to -60 mV were 30.2 and 6.1 ms with and without the preconditioning large depolarization, respectively. Figure 1C shows a histogram of channel open time during repolarization to -60 mV. The channel open time was significantly longer after the large conditioning depolarization. Without a preconditioning depolarization, only 10 channels were open longer than 30 ms, compared with 48 channels after the conditioning depolarization. In addition, after the conditioning depolarization, fewer channels had open times of less than 1.5 ms (including null open traces). Even if the test potential was changed from +40 mV to +20 and 0 mV, and/or the repolarization potential was changed from -60 to -40 mV, the channels tended to open for a longer period of time after large depolarizations.

Figure 1. Cell-attached recordings of smooth muscle α₁-subunits.

Patch pipettes contained 100 mM Ba²⁺ and 2 μM Bay K 8644. Cells were superfused with a high-K⁺, nominally Ca²⁺-free solution. A holding potential of -80 mV was applied. In A, current traces obtained with a paired pulse protocol are shown: a, after the test step (to +40 mV, 7.5 ms) was applied, deactivation of the unitary Ca²⁺ channel current was observed during the subsequent repolarization to -60 mV; b, the test step was preceded by a large conditioning depolarization (to +80 mV, 4 s). The last 7.5 ms of the conditioning depolarization are shown in the figure. The current traces in a and b were obtained from the same cell. B, sum of unitary channel currents obtained from 27 paired traces (nine cells, null traces not included). In the experiments shown in B, test steps of 7.5 or 15 ms depolarization to +40 ms were applied, indicated by the dashed line in the upper voltage protocol trace; the summed channel currents taken from the middle of the test step are shown below. In order to minimize artificial deformity (e.g. changes in the shape of the capacitive surge) not all current traces were used. The decay of the current trace during the repolarization step was fitted with a single exponential function (continuous line). C, histogram of channel open time during the repolarization step (at -60 mV). The histogram includes data obtained by applying the test steps (to +40 mV) for both 7.5 and 15 ms. Filled and open columns correspond, respectively, to data obtained without and with a preconditioning depolarization (to +80 mV, 4 s). The current recordings were analysed only when at least one channel opening was recognized while the paired pulse protocol was sequentially applied. In total, 163 and 153 channels were observed with and without the preconditioning depolarization, respectively. Each null open trace was counted as one channel opening for less than 1.5 ms.

Figure 1B shows the sum of unitary channel currents obtained with (b) and without (a) the preconditioning depolarization. The summed unitary channel currents were constructed from data obtained by applying the test steps (to +40 mV) for 7.5 or 15 ms. The summed tail currents were fitted with a single exponential function. The decay time constant was significantly larger after the large depolarization: 6.7 ms without and 27.4 ms with a large conditioning depolarization.

In three cells, the properties of long channel opening were further examined by changing the duration of the conditioning depolarization (to +80 mV) from 0.1 to 1 and 4 s. Each duration of the large depolarization was applied 58 times. As the duration increased, the number of channels with an open time > 1.5 ms increased from 63 (after 0.1 s) to 83 (after 1 s) and 80 (after 4 s) (Fig. 2A). The mean open times were 16.5, 30.1 and 29.7 ms after 0.1, 1 and 4 s conditioning depolarizations, respectively (Fig. 2B). These results suggest that it takes ∼1 s for most Ca²⁺ channels to enter the second open state, which agrees well with our previous experiments carried out in native smooth muscle cells (Nakayama & Brading, 1993a, 1995a).

Figure 2. Effects of changing the duration of the preconditioning step from 0.1 to 1 and 4 s.

In three membrane patches, the set of three conditioning durations (0.1 s, ▪; 1 s, □; 4 s, ) was repeated 58 times in total. The number of open channels observed is summarized in A. In B, the mean open time (excluding the channels with open times < 1.5 ms) is plotted against the preconditioning duration.

Voltage dependence of channel closure upon repolarization

The effects of changing the repolarizing potential were examined in the experiments illustrated in Fig. 3. After a large conditioning depolarization, one of three different repolarizing potentials (-40, -60 and -80 mV) was applied following a test step (to +40 mV). Each pulse sequence was applied 29 times to the same cell at 30 s intervals. The current traces shown in Fig. 3A are examples obtained at -40 (a), -60 (b) and -80 mV (c). The channel closure during repolarization was slowed by decreasing the negativity of the repolarization step. In this set of pulse sequences, the number of channels observed was similar: 21-23 channels between -40 and -80 mV. When short open times (< 1.5 ms) were excluded, the mean open times during repolarization were 39.5 ms at -40 mV, 21.4 ms at -60 mV and 18.5 ms at -80 mV. Figure 3B shows the sum of the unitary tail currents obtained with these three repolarizing steps. When the decay of the summed tail current was fitted with a single exponential function, the time constants were 87.9 ms at -40 mV, 33.6 ms at -60 mV and 21.6 ms at -80 mV. In four other cells examined, the channel open time consistently decreased with increasing negativity of the repolarization step: the overall mean open times were 29.2 ± 10.8 ms at -40 mV, 20.4 ± 9.1 ms at -60 mV and 13.7 ± 4.1 ms at -80 mV (n = 5). This voltage-dependent change in the deactivation time constant also agrees well with previous observations in native smooth muscle Ca²⁺ channels (Nakayama & Brading, 1993a).

Figure 3. Effects of changing the negativity of the repolarization step.

After a large conditioning depolarization (to +80 mV, 4 s), one of the three repolarizing potentials (-40, -60 and -80 mV) was applied. Each pulse sequence was applied at 30 s intervals. The current traces shown in A are examples obtained at -40 (a), -60 (b) and -80 mV (c). B, sum of unitary tail currents. Each summed trace was constructed with 18-19 traces. All current traces were obtained from the same cell. The decay of the summed tail current was fitted with a single exponential function.

I-V relationship of the unitary Ca²⁺ channel current

Using the characteristic of long channel opening after a large depolarization, the unitary current-voltage (I-V) relationship of the α_1C-b subunit was directly measured. In Fig. 4A, a conditioning depolarization (to +80 mV, 4 s) followed by a ramp step (from +80 to -80 mV, 40 ms) was repeated. The current trace a does not contain channel openings, while the trace b contains only one channel which is continuously open (upper traces). The subtraction current (current a subtracted from current b; lower trace) corresponds to the voltage dependence of the single channel current amplitude between +80 and -80 mV. The I-V relationship obtained from this subtraction is plotted in Fig. 4B. When the data points from -10 to -80 mV were used for least squares fitting, the slope conductance (continuous line) was 25.4 pS. In six cells, the mean slope conductance obtained using the same ramp pulse procedure was 25.9 ± 1.9 pS. This value agrees well with previous estimations obtained for this Ca²⁺ channel subunit without preconditioning depolarizations (Bosse et al. 1992; Gollasch et al. 1996) and for L-type Ca²⁺ channels in smooth muscle cells (Benham et al. 1987; Yoshino et al. 1989; Nakayama & Brading, 1996).

Figure 4. Unitary channel currents (A) and I-V relationship (B) obtained by applying ramp pulses.

A ramp pulse was applied after a large conditioning depolarization (to +80 mV, 4 s). Only the last 5 ms of the preconditioning depolarization is shown. The upper traces in A show three recordings of Ca²⁺ channel currents. The lower trace shows the difference trace obtained when current a is subtracted from b. Based on this subtraction, the I-V relationship of the unitary Ca²⁺ channel current (100 mM Ba²⁺ in the pipette) is plotted in B.

The current trace shown in Fig. 4Ac contains at least three opening channels at the beginning of the repolarization voltage ramp. The channels successively closed as the membrane potential was repolarized. This result indicates that the long channel opening seen during the repolarizing steps after a large depolarization is not due to recovery from the inactivation state, but rather that the channels are open during the large depolarization: Ca²⁺ channels do not or only very slowly inactivate in the second open state (Nakayama & Brading, 1993b).

Coexpression of α- and β-subunits

Paired pulse protocols were also applied to CHO cells in which the cloned smooth muscle α₁-subunit (α_1C-b) and skeletal muscle β-subunit (β_1a) were coexpressed (Fig. 5). In some of the membrane patches, channel opening was prominently reduced after large conditioning depolarizations. Figure 5A shows such an example. In order to demonstrate more clearly the effect of the β-subunit on the large conditioning depolarization, a repolarization step to -40 mV was used: at -40 mV the channel open time was prolonged compared to that at -60 mV. A similar large reduction (by more than 40 %) in the number of channels observed after a preconditioning depolarization (4 s) was seen in 10 out of 18 membrane patches. The results obtained from the 10 patches are summarized in Fig. 5B. When a single test depolarization to +40 mV was applied 104 times, 116 channels were counted (with open times > 1.5 ms). The number of channels observed (with open times > 1.5 ms) was reduced to only 19 after a conditioning depolarization was paired with the test pulse. The mean open time was, however, longer when the large conditioning depolarization was applied: 22.3 ms with and 8.9 ms without the preconditioning depolarization.

Figure 5. Cell-attached recordings of smooth muscle α₁-subunits coexpressed with skeletal muscle β-subunits.

A, current traces obtained from the same cell with a paired pulse protocol. The protocol used in this figure is the same as that shown in Fig. 1A, except for the repolarizing potential (-40 mV). B, histogram showing the channel open time measured during the repolarization step (based on 104 paired traces obtained from 10 cells). In this analysis the data obtained from 10 out of 18 membrane patches was examined. In these 10 membrane patches a large conditioning depolarization reduced the number of open channels by more than 40 %. The details for plotting the histogram are the same as in Fig. 1C.

Figure 6 shows the effect of changing the duration of the preconditioning depolarization from 4 to 0.1 s in the 10 individual membrane patches which showed a large reduction in the number of channels observed after a preconditioning depolarization for 4 s. The ratio of the number of channels observed with to that without a large preconditioning depolarization (NC(pd+)/(NC(pd-)) was plotted against the duration of the depolarization. When a 4 s large preconditioning depolarization was applied, the ratio NC(pd+)/NC(pd-) ranged between 0 and 0.55, with a mean of 0.18. In six out of these 10 membrane patches, the effects of reducing the duration of the depolarization to 0.1 s were examined. The ratio increased in all six patches (mean value 1.14). This recovery of channel opening may be explained by the fast inactivating effect of the coexpressed β_1a subunit (Hullin et al. 1992).

Figure 6. Effects of decreasing the duration of the large conditioning depolarization (to +80 mV) in the 10 membrane patches used to construct the histogram shown in Fig. 5B.

α_1C-b and β_1a subunits were coexpressed in the CHO cells used. These membrane patches showed a reduction in channel opening (number of open channels: NC) after a large conditioning depolarization for 4 s. ○, the ratio of the number of channels observed in the 10 individual membrane patches with (pd+) to that without (pd-) a large conditioning depolarization for 4 s (NC(pd+)/NC(pd-)). , the ratio increased on decreasing the duration of the depolarization to 0.1 s. The number of open channels observed after a 0.1 s preconditioning depolarization was normalized to the number of sweeps applied: the same number of depolarizations with and without a 4 s preconditioning step were applied, but the number of depolarizations with a 0.1 s preconditioning step was lower in some patches.

On the other hand, in the rest of membrane patches examined (8 out of 18), the inactivating effect of the long, large depolarizing step was unclear, despite coexpression of the β-subunit. This can be explained if the α₁- and β-subunits do not assemble properly even though both subunits are expressed. It has recently been reported that a small population of the α-subunits of the maxi Ca²⁺-activated K⁺ channels in coronary smooth muscle cells lacks functional coupling with β-subunits (Tanaka et al. 1997). Another possibility is indicated by the low percentage of open channel traces recorded without conditioning depolarizations (○ in Fig. 6): basic conditions might be different between Ca²⁺ channels showing high and low percentages of open channel traces.

DISCUSSION

The unitary current traces shown in Fig. 1A demonstrate that large preconditioning depolarizations (to +80 mV) of long duration (4 s) do not inactivate smooth muscle α₁-subunits (α_1C-b) expressed in CHO cells. Furthermore, the deactivation time course is much slower after large depolarizations.

The present experiments were carried out using patch pipettes containing 2 μM Bay K 8644, which would produce a maximal Ca²⁺ agonist binding effect (mode 2 gating), a well-known phenomenon resulting in long channel opening (Hess et al. 1984). Bay K 8644 (2 μM) has also been reported to produce maximal kinetic change (a hyperpolarizing shift of the activation curve) for activation of the α_1C-b subunit (Welling et al. 1993). Thus, the prolongation of channel closure after large depolarizations seen in the present study is presumably due to a mechanism other than the Ca²⁺ agonist binding effect. This is consistent with our previous conclusion based on data obtained from native smooth muscle L-type Ca²⁺ channels (Nakayama & Brading, 1995a, 1996). In addition, the change in the deactivation time constants of the summed unitary currents recorded after large conditioning depolarizations seen in the present study is comparable to changes observed in previous whole-cell and single channel recordings of native smooth muscle Ca²⁺ channels (Nakayama & Brading, 1995a, 1996): the values for τ_d-60 (deactivation time constant at -60 mV) of the normal (O₁*) and second open states (O₂*) with the Ca²⁺ agonist bound (the asterisks indicate Ca²⁺ agonist binding) were 3-5 and 30-50 ms, respectively. The present results obtained in cloned Ca²⁺ channels, therefore, reinforce the observation that the slow tail currents induced after large depolarizations seen in smooth muscle cells (in whole-cell mode) are not due to multiple types of Ca²⁺ channels (Nakayama & Brading, 1996). Recently, a similar large depolarization-induced long channel opening phenomenon has been reported in the presence of a Ca²⁺ channel agonist in the L-type Ca²⁺ channels of a rat pituitary cell line (Fass & Levitan, 1996b).

When the voltage of the repolarization step was changed from -80 to -40 mV (in the present study), the deactivation time course was consistently slowed (Fig. 3). This result suggests that the slow deactivation due to the second open state (O₂ and O₂*) involves a charge movement (which would correspond to a voltage-dependent structural change of Ca²⁺ channels). In voltage-clamped cut skeletal muscle fibres, an additional intramembranous charge which moves at high voltage (movement of one-quarter of the total mobile charge with a half-maximal potential ≡+11 mV) has been found (Shirokova et al. 1995). Such charge movement might underlie a transition of the Ca²⁺ channel conformation to the second open state.

As described above, native Ca²⁺ channels in guinea-pig detrusor cells showed essentially the same slow deactivating tail currents as seen in the present study (Nakayama & Brading, 1995a, 1996). In guinea-pig detrusor smooth muscle cells, the slow deactivation due to the second open state is reversibly produced by a large depolarization, even in the presence of a non-hydrolysable ATP analogue (1 mM ATP-γ-S) in the pipette or a high concentration (100 μM) of H-7, which would suppress a broad range of kinase activities, in the bathing solutions (Nakayama et al. 1996; Smith et al. 1999). Recently, similar pharmacological effects on the slow deactivating tail current have been reported in native L-type Ca²⁺ channels in thalamic neurones (Kammermeier & Jones, 1998). Previously in cloned smooth muscle (α_1C-b subunit, from rabbit lung) and cardiac muscle L-type Ca²⁺ channels (α_1C-a subunit, from human tissue), the amplitude of the Ca²⁺ channel current was increased after a large preconditioning depolarization of short duration, even in the presence of phosphorylation inhibitors (Kleppisch et al. 1994; Eisfeld et al. 1996); however, this voltage-dependent facilitation of the Ca²⁺ channel current seen during test steps was not accompanied by a slow deactivating tail current. On the other hand, it has also been reported that a cAMP-dependent protein kinase underlies the voltage-dependent facilitation of rabbit cardiac α₁-subunit (α_1C-a) activation (Sculptoreanu et al. 1993). These reports, together with the results of the present study, suggest that there are numerous mechanisms for prepulse facilitation, even within the α_1C subclass of α₁-subunits.

In our previous study on smooth muscle L-type Ca²⁺ channels, it was also shown that inactivation of the Ca²⁺ channel current is divided into Ca²⁺- (divalent cation-) and voltage-dependent mechanisms, and that these two mechanisms seem to operate separately (Nakayama & Brading, 1993b). In the present study we report that Ca²⁺ channels in the second open state do not (or only very slowly) inactivate during large conditioning depolarizations. This could be more precisely expressed: Ca²⁺ channels in the second open state are resistant to voltage-dependent inactivation, and the influx of divalent cations through Ca²⁺ channels is negligible (consequently divalent cation-dependent inactivation is negligible) during large depolarization steps, because the electrochemical gradient is very small at such membrane potentials. Indeed, when the intracellular Ca²⁺ concentration was increased by releasing Ca²⁺ from the intracellular stores during large conditioning depolarizations (to +80 mV), the Ca²⁺ channel current evoked by the subsequent test step was suppressed (presumably due to Ca²⁺-dependent inactivation of Ca²⁺ channels in the second open state; Nakayama, 1993).

Coexpression of a skeletal muscle β-subunit with most of the known α₁-subunits has been reported to amplify the Ca²⁺ current (increase the current density) evoked by simple rectangular depolarizations (e.g. Hullin et al. 1992; Nishimura et al. 1993). However, in the present study, we showed that in a large population of CHO cells (cell-attached membrane patches), the skeletal muscle β-subunit suppressed the slow tail current formation after a large conditioning depolarization (to +80 mV, 4 s). This paradoxical effect of the skeletal β-subunit could be explained by its fast inactivating effect: the skeletal β-subunit (β_1a) speeds up Ca²⁺ channel inactivation upon depolarizations of ordinary amplitude, despite an increase in the peak inward current (Welling et al. 1993). This explanation is supported by the fact that when skeletal muscle β-subunits were coexpressed with α₁-subunits, the number of open channels decreased, but the mean channel open time was prolonged the by the large conditioning depolarization. In other words, the transition from the normal to the second open state occurred during the large depolarization, but fast inactivation due to the skeletal β-subunit reduced the number open channels.

Since, at the potentials used during the large conditioning depolarization, influx of divalent cations through Ca²⁺ channels is considered to be negligible (because of the small electrochemical gradient), the divalent cation-dependent inactivation mechanism in cloned Ca²⁺ channels (Ferreira et al. 1997; Noceti et al. 1998) does not seem to be a major mechanism in the inactivation seen in our experiments. The inactivating effect of the skeletal muscle β-subunit seen during the large depolarization is thus attributed to modulation of voltage-dependent inactivation.

It has recently been reported that the α₁-subunit of skeletal muscle L-type Ca²⁺ channel (α_1S), coexpressed with skeletal muscle β (β_1b) and α₂δ subunits, shows slow deactivation after a large depolarization of short duration (5-75 ms) (Johnson et al. 1997). The present study provided a similar result: slow deactivation of the tail current in cloned smooth muscle α₁-subunits (α_1C-b) coexpressed with skeletal muscle β-subunits (β_1a) was observed after large depolarizations for 100 ms, although the mean open time was shorter compared to that seen after 4 s large preconditioning depolarizations. In cloned skeletal muscle L-type Ca²⁺ channels, protein kinase A seems to underlie the voltage-dependent slow gating mechanism. However, native smooth muscle (L-type) Ca²⁺ channels, which show time- and voltage-dependent facilitation characteristics most similar to those observed in the present study, do not seem to involve phosphorylation-related mechanisms to develop the second open state during large depolarizations (Smith et al. 1999). Again, it can be deduced that there may be numerous mechanisms underlying voltage-dependent modulation of various types of Ca²⁺ channels, even with respect to tail current formation.

Provided that the conversion of Ca²⁺ channel conformation to the second open state can be modulated, the extent to which the second open state contributes to Ca²⁺ influx will vary, even though the same membrane potential change occurs. GH₃ clonal pituitary cells show Ca²⁺ tail currents with fast and slow time constants. It has been reported that thyrotropin-releasing hormone selectively suppressed the slow component of the Ca²⁺ tail current (Fass & Levitan, 1996a). The present experiments showed that coexpression of a β-subunit (β_1a) modulated the development of the second open state during large depolarizations, which may be due to the high inactivation rate of β_1a. Various inactivation rates have been found among different types of β-subunits (Hullin et al. 1992). It is thus speculated that the heterogeneity of β-subunit expression among tissues possibly produces diversity in the regulation of Ca²⁺ influx via the formation of the second open state. More recently, it has also been revealed that coexpression of α₁- and β-subunits affects the inactivation properties of cloned L-type Ca²⁺ channels, even though the two subunits are not physically coupled on the plasma membrane (Gerster et al. 1999). This new mechanism would provide further complicated and diverse control mechanisms for Ca²⁺ channel kinetics via the β-subunit.

In conclusion, the present unitary current analyses revealed that cloned smooth muscle α₁-subunits expressed in CHO cells preserve the characteristic features of native L-type Ca²⁺ channels seen in smooth muscle cells, namely that large, long-duration depolarizations transform the Ca²⁺ channel conformation from a normal to a long open state (O₂) that is distinct from the Ca²⁺ channel agonist-induced gating (mode 2 gating), and that Ca²⁺ channels in this open state do not or only very slowly inactivate during depolarization and deactivate slowly upon repolarization.