INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE INACTIVATION OF Na⁺ and various voltage-dependent K⁺ channels is determined by both fast and slow mechanisms (16, 49). During depolarization, Na⁺ current is inactivated by a fast Hodgkin-Huxley-type mechanism within 20 ms. If the depolarizing pulse is prolonged, the recovery of the Na⁺ current from inactivation is much delayed during repolarization, suggesting an additional slow-inactivation mechanism (37). Recently, the structural basis for the slow inactivation was explored by mutagenesis studies of the -subunit, and separate structures were suggested for the fast and slow kinetics, respectively (2, 32).

The -subunit of the L-type Ca²⁺ channel has very similar structure to the Na⁺ channel, and a slow inactivation has also been reported (3, 12, 22, 30). Namely, prolongation of the depolarizing pulse for several tens of seconds results in a very slow recovery of repolarization, although the decay of L-type Ca²⁺ current (I_Ca,L) is obviously completed within a 200-ms depolarization (3). The fast inactivation has been explained by both the Ca²⁺-mediated and voltage-dependent mechanisms (8, 21, 26, 28). However, the nature of the slow inactivation of the L-type Ca²⁺ channel has still not been clarified.

Recently, we noticed a seasonal variation in the slow inactivation of I_Ca,L when the inactivation kinetics of the L-type Ca²⁺ channel were compared throughout the year. The slow inactivation was evident when the myocytes were prepared from guinea pigs bred at ambient temperature in winter. Given this observation, the present study aims to test our hypothesis that the slow inactivation of the Ca²⁺ channel is induced by cold acclimation of the experimental animals.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Preparation. The experimental animals were purchased from a breeding farm that keeps guinea pigs at room temperature without air conditioning in Nakatsu-City Oita Prefecture, Japan. According to the Japan Weather Association, the daily mean of the ambient temperature of Nakatsu-City varied seasonally between 4.1 to 29.1°C in 1995; the experiments in the present study were conducted in 1995-1996.

The guinea pigs were kept at room temperature (~25°C) in our laboratory and used generally within 1 wk after obtaining them from the breeder. In several experiments, guinea pigs were kept for >3 wk in a cold room, in which the temperature was kept constant at 5°C before they were used. During the period of cold exposure the body weight of guinea pigs increased from ~200-250 to ~350-450 g, similar to control animals kept at room temperature.

Single myocytes were dissociated from left ventricular myocardium as described previously (18, 35). In brief, under deep anesthesia (intraperitoneal injection of pentobarbital sodium, 50 mg/kg), the ascending aorta was cannulated in situ. While the coronary perfusion was maintained with the control Tyrode solution, the heart was dissected out and fixed on the Langendorff-type perfusion system at 37°C. Then the perfusate was switched to the nominally Ca²⁺-free solution until the heartbeat stopped. The collagenase solution [Sigma Chemical, St. Louis, MO; type 1, 40 mg/100 ml low Ca²⁺ (30-40 µM) solution] was perfused for ~20 min. Finally, the enzyme was washed out by perfusing a high-K⁺/low-Cl solution. Within the same solution the myocytes were mechanically dissociated and stored in a cell culture medium (Eagle's medium, Flow Laboratories, McLean, VA) at room temperature.

Solutions used for the cell isolation. The composition of the control Tyrode solution used for cell isolation was (in mM) 140 NaCl, 5.4 KCl, 1.8 CaCl₂, 0.5 MgCl₂, 0.3 NaH₂PO₄, 5.5 glucose, and 5 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES); the pH was adjusted to 7.4 with NaOH. Ca²⁺-free Tyrode solution was prepared by omitting CaCl₂ from the control Tyrode solution. The high-K⁺/low-Cl solution contained (in mM) 70 glutamic acid, 25 KCl, 10 taurine, 10 KH₂PO₄, 11 glucose, 0.5 ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), and 10 HEPES (pH 7.3 with KOH).

Solutions used for recording the I_Ca,L. The composition of the external solution for recording the I_Ca,L was the same as the control Tyrode solution, except that the external K⁺ was substituted with equimolar Cs⁺. The pipette solution contained (in mM) 110 Cs-aspartate, 5 Mg-ATP, 5 EGTA, 20 CsCl, 0.2 GTP, 10 tetraethylammonium-Cl, and 5 HEPES; the pH was adjusted to 7.2 with CsOH. Five millimolar guanosine 5'-O-(3-thiotriphosphate) (GTPS) (Sigma) or 5 mM guanosine 5'-O-(2-thiodiphosphate) (GDPS; Sigma) was added to the control pipette solution when appropriate.

Pertussis toxin incubation. In some experiments, the dissociated myocytes were incubated at 37°C in Eagle's medium containing pertussis toxin (PTX; 5 µg/ml, Sigma) for more than 5 h. Block of the GTP-binding protein (G_i) was confirmed by examining the antagonistic action of acetylcholine (ACh) against the isoprenaline-induced increase of I_Ca,L.

Voltage-clamp experiments. I_Ca,L was measured by the whole cell patch clamp technique (9) using an Axopatch-1D amplifier (Axon Instruments, Foster, CA). The glass suction pipette had a tip diameter of ~4 µM and a resistance of 2-3 M when filled with the control internal solution. After the formation of a giga-ohm seal, a strong suction was briefly applied to the inside of the pipette to rupture the patch membrane. The liquid junction potential between the pipette solution and the external solution (10 mV) was corrected for all membrane potential recordings. All experiments were carried out at 35 ± 0.5°C. Holding potential for recording I_Ca,L was 40 or 100 mV. The voltage dependence of inactivation was measured using a double-pulse protocol. The conditioning pulse and test pulse were separated by a 40-ms gap at 40 mV during which both tetrodotoxin-sensitive Na⁺ current and T-type Ca²⁺ current were inactivated (14).

The exchange of the bath solution was complete within 15 s after switching perfusates at the inlet of the chamber. The current and voltage signals were recorded on a digital magnetic tape (RD101, TEAC, Tokyo, Japan) and played back later for computer analysis.

Statistics. Values are expressed as means ± SD. Statistical difference was examined with paired or unpaired Student's t-test (P < 0.05).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Seasonal variation in the voltage dependence of I_Ca,L inactivation. The voltage dependence of I_Ca,L inactivation was examined by applying the pulse protocol shown at the top of Fig. 1 every 7 s. The membrane potential was first stepped for 400 ms (conditioning pulse), and the degree of inactivation during the conditioning pulse was measured as the relative amplitude of I_Ca,L activated by the following test pulse to 0 mV. When the inactivation of I_Ca,L was removed step by step by setting the conditioning pulse to more negative potentials, the peak amplitude of I_Ca,L induced by the following test pulse was increased. This effect of increasing I_Ca,L usually saturated at potentials more negative than 40 mV (Fig. 1Aa). We found, however, that the I_Ca,L amplitude increased in some experiments as the conditioning pulse was made more negative than 50 mV as shown in Fig. 1A, b and c. To determine the "inactivation curve," the amplitude of I_Ca,L was determined as a difference of current levels at the peak and at the end of the test pulse, and the relative amplitude of I_Ca,L was plotted against the membrane potential during the conditioning pulse (Fig. 1B). The dynamic range of inactivation was between 50 and 10 mV in curve a, whereas it was observed at more negative potential range in curves b and c. These inactivation-voltage relationships remained stable in a given cell throughout the experiment (<20 min).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Inactivation curve of L-type Ca2+ current (ICa,L) determined by conventional pulse protocol. Original current recordings (A) were obtained with the test pulses to 0 mV, which followed the 400-ms conditioning pulse to various levels as indicated. Test pulse was separated from the conditioning pulse by 40 ms. In B, curves are the fit of equation 1. Data in a were obtained in summer, whereas, b and c were obtained in winter. Dashed lines in A indicate current zero level.

For the limited purpose of comparing these inactivation curves, the degree of inactivation (f) was expressed as a function of the membrane potential (V_m) with a modified Boltzmann equation, which includes a noninactivating component (A)

(1)

where V_half is the potential one-half inactivation, and s is the slope factor. The values of parameters obtained by fitting equation 1 to curve a were V_half = 31.5 mV, s = 5.1 mV, and A = 0.23. The parameters for curve b were V_half = 49.8 mV, s = 11.1 mV, and A = 0.08, and for curve c were V_half = 73.2 mV, s = 21.6 mV, and A = 0.01.

The variation in the inactivation curve depended on the season; typically curves b and c in Fig. 1 were obtained in winter, whereas curve a was obtained in summer. The values of V_half and s obtained between May and September are roughly comparable with those described in the literature (1, 14, 17, 19, 24, 42). The average values obtained from 83 experiments during this period were V_half = 34.9 ± 7.9 mV and s = 6.4 ± 2.2 mV. In experiments between January and April, the values of V_half were significantly more negative (negative shift of the inactivation curve), and s values were significantly larger than in May through September (P < 0.05, unpaired Student's t-test).

To examine the temperature dependence of inactivation, the values of V_half and s were plotted in Fig. 2, A and B, against the ambient temperature (Fig. 2, top). The relationships were fitted with regression lines, with coefficients of 0.77 and 0.62 for Fig. 2, A and B, respectively. Thus it seems that the voltage-dependence of I_Ca,L inactivation might vary with the seasonal changes in the ambient temperature.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Relationship between the potential of half inactivation (Vhalf) or slope factor (s) and temperature. Daily mean of the ambient temperature in 1995 at the place of animal breeding farm is shown at top. Vhalf  and s were obtained by fitting equation 1 to the experimental inactivation curve in each experiment and were plotted against the ambient temperature in A and B, respectively. Regression line is the fit to all data points.

Induction of the negative shift of I_Ca,L inactivation by cold acclimation. To further examine the cause and effect relationships between the temperatures to which the experimental animals were exposed and the occurrence of the negative shift of I_Ca,L inactivation, we obtained adult guinea pigs in July-August and divided them into two groups. In the experiments in Fig. 3, one group was kept at 5°C for >3 wk (cold acclimation) before the experiment that measured I_Ca,L and was compared with the control group, which was kept at ~25°C. The inactivation curve obtained from cold-exposed guinea pigs consistently showed V_half to be more negative than 50 mV, with a mean of 76.6 ± 6.9 mV (n = 10), whereas V_half of control animals was 35.4 ± 9.9 mV (n = 36). The values of s were larger in cold-exposed animals (13.6 ± 3.8 mV) than in control animals (6.6 ± 2.6 mV) (P < 0.05, unpaired Student's t-test).

Slow-gating mechanisms underlying the negative shift of inactivation. To identify the gating steps that were modulated by the cold acclimation, channel gating was compared in myocytes with and without the negative shift of I_Ca,L inactivation. No obvious difference was observed in the time course of I_Ca,L decay during a depolarizing pulse to 0 mV. The major part of I_Ca,L decay was well fitted by a single exponential curve of time constant 14.3 ± 4.7 ms (n = 5) in myocytes, with the negative shift not significantly different from the 12.5 ± 2.5 ms (n = 5) observed in myocytes without the negative shift (P > 0.05, unpaired Student's t-test).

Because inactivation depends on the activation of the channel, as suggested for Na⁺ channels (31), we also examined activation of I_Ca,L. Test pulses of 400-ms duration were given to various potentials from a holding potential of 40 mV. The amplitude of I_Ca,L was determined as a difference of current levels at peak I_Ca,L and at the end of the 400-ms test pulse. On the assumption of a reversal potential of +50 mV for I_Ca,L, the chord conductance of peak I_Ca,L was calculated at various potentials, normalized to the maximum value (d), whose voltage relationship was fitted with the Boltzmann equation Values of V_half and s in cells with and without slow-inactivating kinetics were 11.3 ± 5.1, 3.2 ± 1.5 (n = 5), and 12.5 ± 3.2, 4.3 ± 1.1 mV (n = 5), respectively, and these average values were not significantly different (P > 0.05, unpaired Student's t-test).

These findings suggest that the inactivation of I_Ca,L observed at potentials more negative than 50 mV might be determined by additional slow-inactivation mechanisms (3, 12, 21, 32). In support of this view, the inactivation at potentials more negative than 50 mV could be removed by setting the holding potential more negative. In the experiment shown in Fig. 4, the negative shift of the inactivation curve was first confirmed (V_half = 62.2 mV, s = 12.2 mV) when the conventional holding potential of 40 mV was used. In contrast, after the holding potential was made more negative (100 mV), the double-pulse protocol revealed a typical euthermic inactivation curve (V_half = 31.6 mV and s = 5.4 mV). On average (n = 4), V_half and s were 61.7 ± 0.9 and 10.0 ± 3.6 mV, with the holding potential of 40 mV, and 31.2 ± 0.6 and 5.1 ± 0.6 mV, with the holding potential of 100 mV, respectively.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Comparison of Vhalf and s between myocytes with and without the cold exposure of the animal. Values of Vhalf (A) and s (B) were compared between myocytes dissociated from guinea pigs with and without cold exposure (5°C, cold; 25°C, warm).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Inactivation curves measured with different holding potentials. Negative voltage dependence of the inactivation curve recorded using a holding potential of 40 mV in cold-exposed animals () was removed by hyperpolarizing membrane to 100 mV (). Pulse protocols are shown in inset. Families of currents at top correspond to all data points in the graph.

Slow-inactivation kinetics. The above results indicate that the "inactivation curve" measured in the present study does not reflect the voltage dependency of genuine steady-state inactivation. The slow inactivation induced at the holding potential of 40 mV might not be completely removed by the standard (400 ms) conditioning pulse to 100 mV. Therefore, the time course of removing inactivation was measured using the paired-pulse protocol shown in Fig. 5, top, where I_Ca,L was first inactivated by depolarizing the membrane to 0 mV for 300 ms from a holding potential of 40 mV. Then the inactivation was removed for various periods at different conditioning potentials before applying the test pulse. The normalized amplitude of I_Ca,L at the test pulse was plotted against the conditioning pulse duration in Fig. 5, and thereby the time courses of recovery were compared between cells with (Fig. 5B) and without (Fig. 5A) the negative shift of inactivation. The curve superimposed on the recovery time course in Fig. 5 is the fit of a single exponential (Fig. 5A) or a sum of two exponential components (Fig. 5B). In the experiment in Fig. 5A, the recovery was single exponential with time constants of 10 and 87 ms at 100 (a) and 40 mV (b), respectively. In contrast, removal of inactivation was double exponential and was slow in the cell shown in Fig. 5B. Time constants of the slow component were 2,217 and 4,087 ms at 80 (Fig. 5Ba) and 50 mV (Fig. 5Bb), respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Time course of the removal of inactivation in animals with and without cold exposure. Time course of the removal of inactivation was measured using myocytes of guinea pigs bred without (A) and with cold exposure (B). Pulse protocol is shown at top. All data points were fitted by an exponential in A, whereas a sum of exponentials was fitted in B. Numbers in graph indicate conditioning potential or the time constant of each exponential curve fitted. In B, the time course was demonstrated at the slower (a, b) and faster time base (a', b'). Curves apart from data points in a' and b' indicate the slower component. Vhalf and s were 31.3 and 5.2 mV in the myocyte in A and 57.8 and 15.6 mV in B, respectively.

The time course of development of the slow inactivation was measured using the pulse protocol shown in Fig. 6, top. The slow inactivation of I_Ca,L was removed by setting the holding potential at 100 mV, and conditioning depolarizations of various duration were applied immediately before the constant test pulse to 0 mV. The normalized amplitude of I_Ca,L at the test pulse was plotted against the conditioning pulse duration in Fig. 6. Superimposed curves are the fit of a sum of two exponentials. In the measurements shown in Fig. 6, time constants of inactivation at 40 mV (a, a') were 160 and 4,532 ms, and those at 30 mV (b, b') were 64 and 2,796 ms. Slow inactivation was consistently observed in cells with V_half more negative than 50 mV. On average, the time constant of slow inactivation was 6.8 ± 3.0 s (n = 6) at 40 mV.

View larger version (25K): [in this window] [in a new window]   Fig. 6.   Slow development of the inactivation in the cold-exposed guinea pig. Double-pulse protocol for measuring the inactivation time course is shown at top. Slow inactivation of ICa,L was removed by using a holding potential of 100 mV and then allowed to inactivate at a given potential for various periods. Initial parts of a and b are shown with expanded time scale in a' and b', respectively. Numbers in the graph indicate the conditioning potential or the time constant of the exponential curve fitted. Note that the inactivation does not develop at 40 mV in control guinea pigs without cold exposure.

Time constants of inactivation or removal of inactivation were plotted against potentials on a logarithmic scale in Fig. 7; data points in Fig. 7, A and B, were from cells without and with the negative shift of inactivation, respectively. In both groups time constants of the fast component were <100 ms and showed a slight voltage dependence, as previously reported (4, 5, 7, 8, 15, 45). The time constants of the slow component in Fig. 7B were >2 s, with little voltage dependence. If the time constants of the faster component in Fig. 7B were compared with those in Fig. 7A, the former were always several-fold larger than the latter. It is suggested that the rapid reactivating kinetics might also be modified in cells with slow inactivation.

View larger version (6K): [in this window] [in a new window]   Fig. 7.   Voltage dependence of the time constant of the recovery from inactivation. Time constant(s) of the removal of inactivation in control and cold-exposed animals were plotted against potentials on a semi-logarithmic scale. A: control. B: cold-exposed guinea pigs.  in B indicate time constants for the slow component. See text for detailed explanation.

Molecular mechanisms: participation of PTX-insensitive G protein. To clarify the molecular mechanisms of slow inactivation, various factors were examined that are known to modify the cardiac I_Ca,L. It has been reported that the cardiac L-type Ca²⁺ channels are modulated by G proteins indirectly (10) or directly (34, 47). Indirect G protein effects in turn occur by phosphorylation of the channel by either protein kinase A (PKA) (20) or protein kinase C (PKC) (27, 38). First, we tested the involvement of these kinases in the slow inactivation. Neither 2 µM staurosporine nor 100 µM 1-(5-isoquinolinylsulfonyl)-2-methylpiperazine reversed the negative shift of the inactivation curve (n = 5), excluding involvement of PKC, PKA, or Ca²⁺-calmodulin kinase (48).

The involvement of G protein was tested by internally dialyzing myocytes with 5 mM GDPS, a selective blocker of ubiquitous GTP-binding proteins. In the experiment shown in Fig. 8, the inactivation curve, measured soon after rupturing the patch membrane, showed the negative shift. As the internal dialysis progressed, the inactivation at very negative potentials was removed and finally a normal inactivation curve was obtained 5 min after starting the internal dialysis. In seven cells with negative shift of inactivation, V_half and s obtained 5 min after the cell dialysis with GDPS were 26.7 ± 5.4 and 5.7 ± 2.7 mV, respectively, which were significantly different from those measured immediately after the start of experiment (P < 0.05, V_half = 72.7 ± 13.5 mV and s = 21.6 ± 8.0 mV, paired Student's t-test).

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Effect of guanosine 5'-O-(2-thiodiphosphate) (GDPS) on the inactivation curve of the cold-exposed guinea pig. GDPS (5 mM) was added to the pipette solution, and the voltage dependence of the inactivation was determined using a myocyte dissociated from the cold-exposed guinea pig. A fraction of the data points indicated the negative voltage-dependent characteristic to the cold-acclimated animals at the beginning of the whole cell recordings (). After 5 min the inactivation at negative potentials was abolished (). Intermittent curve is the fit of equation 1 to  more negative than 30 mV, which gave Vhalf = 85.3 mV and s = 16.7 mV. Parameters for the curve fitted to the  were Vhalf = 22.4 and s = 4.6 mV. Families of currents at top correspond to data points of 20, 30, 50, 70, 90, and 110 mV in the graph.

It seems that the voltage dependence of inactivation obtained at the beginning of experiments consisted of two Boltzmann components. It may be speculated that a part of I_Ca,L channels was relieved from the slow inactivation by a rapid diffusion of GDPS into the cell soon after starting of the whole cell patch clamp recording because these two components were not observed without GDPS in the pipette solution. Provided that GDPS specifically inhibits activity of GTP-binding proteins, these results suggested that the negative voltage shift of inactivation was induced by an unknown mechanism depending on a GTP binding protein.

In an attempt to identify the GTP binding protein underlying the GDPS effect, the dissociated myocytes were incubated at 37°C in culture medium containing PTX (5 µg/ml) for >5 h. Blockage of the GTP-binding protein (G_i) was confirmed by the disappearance of antagonistic action of ACh against the isoprenaline-induced increase of I_Ca,L in the PTX-treated myocytes. However, the negative shift of inactivation was still observed in all six cells.

If the membrane of the cardiac cells of control animals has a specific G protein that is involved in the slow inactivation in cold-acclimated animals, a nonhydrolyzable GTP analog that activates ubiquitous GTP binding proteins might induce the negative shift of inactivation. In the experiment in Fig. 9, 5 mM GTPS was added to internal pipette solution. The inactivation curve measured soon after starting the whole cell recording showed no obvious negative shift. Essentially the same inactivation curve was obtained 10 min after starting the cell dialysis. Transient increase of I_Ca,L was observed soon after starting the cell dialysis with GTPS, which was followed by a rundown as reported (25). In five experiments, V_half and s of inactivation curves obtained soon after starting whole cell recording were 23.1 ± 10.4 and 6.9 ± 2.6 mV, respectively. V_half and s obtained 5 min after the cell dialysis with GTPS were 24.2 ± 2.3 and 5.0 ± 0.8 mV, respectively, which were not different from those just after the start of experiment (P > 0.05, paired Student's t-test).

View larger version (16K): [in this window] [in a new window]   Fig. 9.   Effect of GTPS on the inactivation curve of the control guinea pig. Perfusion of 5 mM GTPS failed to modify the inactivation of the myocyte of control animals soon after starting the experiment (; Vhalf = 28.7 mV, s = 7.5 mV) and after 10 min (; Vhalf = 26.7 mV, s = 6.0 mV). Families of currents at top correspond to data points of 10, 20, 30, and 50 mV in the graph.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

One of the major findings of the present study is that the quasi-steady-state inactivation curve of I_Ca,L, measured in dissociated cardiac myocytes, can be modulated by exposing the experimental animal to cold temperatures (5°C) for >3 wk (Fig. 3). This finding correlates to the seasonal difference in the inactivation curve observed in myocytes, when the experimental animals were kept without the temperature control in breeding rooms (Fig. 2). Although the molecular mechanisms were not clarified in the present study, an induction of G protein-coupled, slow-inactivation mechanisms by the cold acclimation was strongly suggested. The modulation of I_Ca,L inactivation by the cold acclimation is a novel finding to the best of our knowledge.

The inactivation curve measured in the present study might be explained by assuming fast- and slow-inactivation gates. The fast inactivation was observed in myocytes that were dissociated from animals without the cold exposure. This fast inactivation may be responsible for the I_Ca,L decay during a short depolarizing pulse and correspond to the Ca²⁺-mediated and voltage-mediated inactivation so far described in literature (7, 21, 26, 28). Even in myocytes obtained from the cold-exposed animals, the inactivation curve obtained after removal of the slow inactivation using the holding potential of 100 mV may also correspond to the fast-inactivation mechanisms.

The negative shift of the inactivation curve (Fig. 1) observed in myocytes obtained from the cold-exposed animals can be well explained by assuming the existence of slow-inactivation mechanisms. At the holding potential of 40 mV a large fraction of the channels is inactivated by the slow mechanisms and only a small fraction of I_Ca,L channels might be relieved from the slow inactivation during 400 ms of the conditioning pulse according to equation where S_t gives the fraction of channels at time t, which is not inactivated by the slow mechanisms (S). Because the time constant of inactivation is not obviously dependent on membrane potential, the value of exp(t/) for a 400-ms pulse should remain almost constant (0.8-0.9) at different membrane potentials. Provided that S  S₀, the value of S_t at the end of 0.4-s pulse is approximated as Namely, the value of S_0.4 reflects the relationship of S-V_m. The experimental inactivation curve may also suggest that the S-V_m relationship has a shallower slope compared with the steady-state fast inactivation-voltage (F-V_m) relationship. The variation in the experimental measurements of V_half and s in the present whole cell current recordings might be determined by different fractional compositions of both the fast- and slow-inactivation kinetics in a given individual channel or by different populations of the modified channels within a given cell.

The slow inactivation of I_Ca,L in ventricular myocytes has been studied in connection with staircase phenomenon observed in the developed tension. Namely, the magnitude of contraction initially decreases step by step when a higher stimulation frequency is started, accompanied by a decrease in the amplitude of Ca²⁺ current (30). Although the degree of this slow inactivation was decreased by adding EGTA into the pipette solution for whole cell voltage clamp, the frequency-dependent decrease of I_Ca,L was observed. Because the diastolic period was long enough to allow a complete recovery from the conventional fast-inactivation mechanisms, the existence of slow inactivation was suggested. Boyett et al. (3) analyzed the slow kinetics in detail and described the slow kinetics by the first-order reaction scheme. It was demonstrated that the removal of inactivation is accelerated by hyperpolarizing the membrane. In contrast, we failed to observe an obvious voltage dependency of the time constant of recovery from the slow inactivation (Fig. 5). Furthermore, in the previous study (3) the increase in amplitude of I_Ca,L with hyperpolarization saturated at potentials negative to 50 mV. It might be suggested that the mechanism of slow inactivation assumed in the present study is different from that reported by Boyett et al. (3). However, the different pulse protocols interfere with comparing the slow kinetics in detail between these studies. The time constant of the fast decay of Ca²⁺ current was not different between cells dissociated from control and cold-acclimated guinea pigs, which suggests that Ca²⁺-dependent inactivation is not significantly modified by the cold exposure. It has been suggested that voltage- and Ca²⁺-dependent inactivation are regulated by distinct sites on the _1C-subunit (33). The structural determinants for voltage-dependent inactivation have been attributed to sequences near or in S6 segments of domains I, III, and IV of the ₁-subunit (46, 51). The expression of ₁-subunits is characterized by alternative splicing, which generates multiform channels. However, the physiological implication of this alternative splicing is poorly understood. It has been shown that alternative splicing of the gene encoding the _1C-subunit of L-type Ca²⁺ channels contributes to the differences in the voltage dependence of the sensitivity toward dihydropyridines (DHPs) (40) and to the DHP (44) tissue selectivity. Recent investigation of human class C L-type Ca²⁺ channel revealed diversity of inactivation kinetics among the well-defined _1C,77 and its two splice variants (_1C,72, _1C,86). The decay of Ba²⁺ current (I_Ba), which indicates that the voltage-dependent inactivation alone was 8-10 times faster in _1C,86 than the others, and the inactivation curve of I_Ba through _1C,72, _1C,86 were negatively shifted by 11 and 6 mV (41).

The present study suggests that G protein is involved in the slow inactivation of cardiac L-type Ca²⁺ channels in cold-acclimated guinea pig. Recent studies suggest that G and G bind to the cytoplasmic linker region between transmembrane repeats I and II of _1A, _1B, and _1E (43, 50). Such a direct binding site, however, has not been found in isoforms of L-type Ca²⁺ channel (_1C, _1D, and _1S). The results in the present study may strongly suggest a new possibility that alternative splicing of ₁-subunit that results in G protein-dependent inactivation kinetics might occur also in cardiac L-type Ca²⁺ channels during a course of cold acclimation of animals. The experimental findings in the present study might be explained by assuming that alternative splicing of ₁-subunit that results in G protein-dependent modification of the inactivation kinetics may occur during the course of cold acclimation. This view is consistent with the block of the negative shift of the inactivation curve by GDPS in the cold-acclimated guinea pigs and is also consistent with the finding that GTPS failed to cause the negative shift in control animals. It should be noted, however, that the present study does not exclude a possibility that the G protein effects on slow inactivation are still indirect through a phosphorylation of the channel by an unknown protein kinase that was not inhibited by the blockers used in the present study.

Perspectives

The long-term regulation of the L-type Ca²⁺ channel in heart described in the present study may bring new insight into the regulation of the L-type Ca²⁺ channel. We failed to specify the G protein responsible for the slow inactivation, although PTX-sensitive G protein was excluded. Similarly, we know almost nothing about the chemical background for the induction of the molecular changes responsible for the slow inactivation. Thus systematic studies from various fields are needed to reveal the role of heart in cold acclimation.