INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MG²⁺ is abundant in the cytoplasm and exerts several regulatory functions in cardiac myocytes. Intracellular Mg²⁺ (Mg) acts as a stimulator, often with ATP, whether it is phosphorylation dependent or not (18, 32). Blocking effects of Mg have been also reported (1, 26-29). Pelzer et al. (20) presented both stimulatory and inhibitory effects of Mg on the L-type Ca²⁺ channel in guinea pig ventricular myocytes. How are these Mg²⁺ effects physiologically relevant? As to the Mg²⁺ block, we have previously postulated that cAMP-dependent phosphorylation reduces the sensitivity to Mg²⁺ block so that the L-type Ca²⁺ channel can be activated by -adrenoreceptor stimulation in the heart (29). However, similar blocking effects of Mg²⁺ for mammalian cardiac myocytes have been lacking with the exception of a single brief report (1). Moreover, Hirahara et al. (9) reported that intracellular dialysis with Mg²⁺-depleting solutions produced only a transient and marginal increase of L-type Ca²⁺ channel current (I_Ca) in guinea pig ventricular myocytes. Recently, Pelzer et al. (20) have shown a modest Mg²⁺ blocking effect that disappear in the presence of the nonspecific protein kinase inhibitor K252a. In our hands, the Mg²⁺ block persisted even under the condition that the phosphorylation process was completely inhibited by the depletion intracellular ATP (28). The question arose as to whether the same Mg²⁺-dependent block observed in frog cardiomyocytes exists in the mammalian heart. According to our hypothesis that phosphorylation of the L-type Ca²⁺ channel in frog ventricular myocytes increases I_Ca by relieving Mg²⁺-dependent block of the channel, a similar mechanism may operate in guinea pig ventricular myocytes, because I_Ca of guinea pig ventricular myocytes is regulated by cAMP-dependent phosphorylation (19, 25).

In this study, we employed a method of intracellular dialysis to vary the concentration of Mg²⁺ inside of guinea pig ventricular myocytes. We found that temperature is a key factor controlling the Mg²⁺-dependent regulation of the L-type Ca²⁺ channel in guinea pig as well as frog ventricular myocytes.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell preparation. Single ventricular cells were obtained from either guinea pig or frog hearts by enzymatic dissociation as previously described (8, 22). Animals were used in accordance with the "Guiding Principles for the Care and Use of Animals" approved by the Council of the Physiological Society of Japan. Adult male guinea pigs weighing 200-400 g were anesthetized with pentobarbitone sodium (30 mg/kg ip), and, under artificial respiration, the heart was rapidly removed. Frogs were killed by decapitation, the spinal cord was destroyed, and the heart was then excised. The heart of either species was mounted on a Langendorff apparatus for retrograde perfusion of the aorta. Hearts were perfused (15-20 min at 32°C) with Ca²⁺-free Tyrode solution (guinea pig) or Ca²⁺-free Ringer solution (frog) (see Solutions and chemicals) supplemented with collagenase (0.12 mg/ml) (guinea pig) or with a mixture of Yakult collagenase (0.040 mg/ml, Yakult), Wako collagenase (0.40 mg/ml, Wako Pure Chemical Industries), and type III trypsin (0.06 mg/ml, Sigma) (frog). The hearts were then rinsed with storage solution (see Solutions and chemicals), and cells were dispersed and filtered through a nylon mesh (200 µm). Dispersed cells were collected by centrifugation at 65 g for 1 min, maintained at room temperature for the first hour, and finally stored at 4°C until used in experiments.

Solutions and chemicals. Tyrode solution was used as the standard external solution for guinea pig ventricular myocytes. Ca²⁺-free Tyrode solution contained (in mM) 135 NaCl, 1.0 MgCl₂ · 6H₂O, 5.4 KCl, 0.33 NaH₂PO₄ · 2H₂O, 10.0 HEPES, and 5.5 glucose; pH 7.4 (adjusted with NaOH). To record I_Ca in guinea pig myocytes, Na⁺-free external and internal solutions were used. Na⁺-free external solution was composed of (in mM) 137 N-methyl-D-glucamine (NMDG)-HCl, 20 CsCl, 5 glucose, 10 HEPES, and 2 CaCl₂; pH 7.4 (adjusted with CsOH). In the experiments shown in Fig. 6C, the equimolar NMDG-HCl of the Na⁺-free external solution was replaced with NaCl and TTX (3 µM) was added (100% Na⁺ external solution). Na⁺-free internal solution was composed of (in mM) 136 CsCl, 3.0 tris(hydroxymethyl)aminomethane ATP salt (Tris-ATP), 10 1,2-bis(2-aminophenoxy)- ethane-N,N,N',N'-tetraacetic acid (BAPTA), 5 EDTA, and 10 HEPES; pH 7.2 (adjusted with CsOH).

Temperature control. The temperature of the bath liquid was measured using a platinum resistance thermometer (dimensions, 2.3 × 2.3 × 2.0 mm, type RMB11101, Yamari Industries). Temperature was manually controlled using a Pelletier device (NetsuDenshi Kogyo) that heated or cooled (by means of direct feedback current) the polyethylene tubing (outer diameter, 1.5 mm; length, 10 cm) carrying liquid to the experimental chamber. At the same time, excess heat was drawn from the Pelletier block by coolant water circulated through an attached brass fixture (10 ml in volume). The flow rate of the external perfusate was kept constant at 1 ml/min so that total replacement of the solution in the chamber (200-µl volume) was complete within 2 min. Bath temperature could be well controlled between 9 and 40°C at an ambient temperature of ~24°C.

Electrophysiological recording and data analysis. Whole cell Ca²⁺ currents (I_Ca) for both the guinea pig and frog ventricular myocytes were recorded using a patch-clamp amplifier (Axopatch 200A, Axon Instruments) and filtered through a four-pole Bessel low-pass filter with a cutoff frequency of 5 kHz. Curve fits and data analysis were performed with pCLAMP software (Axon Instruments). Pipette resistance was 1-1.5 M when filled with Na⁺-free internal solution. Membrane capacitance was determined from the current amplitude elicited in response to hyperpolarizing (from a holding potential of 80 to 130 mV for 50 ms) and depolarizing (from 130 to 80 mV for 50 ms) ramp voltage pulses at a rate of 1.0 V/s. Capacitance measurements by the ramp pulses were always made at the beginning and end of each experiment. The averaged cell capacitances used in this study were 74.0 ± 1.6 pF (n = 134) for guinea pig ventricular myocytes and 77.8 ± 5.2 pF (n = 30) for frog ventricular myocytes. In addition, cell capacitance compensation in combination with the series resistance (R_s) compensation (70-80%) of the Axopatch 200A amplifier was used to monitor cell capacitance and R_s throughout the experiments. Separately, in a series of experiments (see Fig. 6), cell capacitance and R_s (2.56 ± 0.08 M, n = 31) were monitored but without using R_s compensation after each protocol of obtaining current-voltage relationships (see below) by applying a voltage ramp as indicated above together with a 20-mV hyperpolarizing voltage step from a holding potential of 80 mV. This enabled us to estimate efficiency of cell dialysis. R_s measured with voltage steps were comparable with R_s readings of the Axopatch 200A amplifier and were usually twice the pipette resistance. Cells were directly transferred from KB medium to the experimental chamber, and current recordings were started after 5- to 10-min perfusions of the external solutions. Approximately 60-70% of guinea pig cells subjected to Na⁺-free external solution were quiescent. Currents other than I_Ca were eliminated almost totally by the use of the internal and external solutions specified above. I_Ca was monitored every 5 s with clamp steps from 40 to +40 mV in 10-mV increments from a holding potential of 80 mV. For guinea pig myocytes, to inactivate outward-going current through TTX-sensitive Na⁺ channels and inward-going current through T-type Ca²⁺ channels, a 200-ms conditioning prepulse to 40 mV was applied just before eliciting I_Ca. I_Ca amplitude was determined as the absolute value of the peak inward current, because leakage currents were usually negligible (otherwise the data were excluded). In guinea pig and frog myocytes, peak I_Ca was always observed at transmembrane potentials between 0 and +20 mV with the 10-mV step pulse protocol described above.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Pronounced temperature sensitivity of I_Ca increased by low-Mg solutions. A sudden change in temperature from 24 to 32°C caused an increase in the amplitude of I_Ca and a shortening of time to peak and inactivation time constants (_inact) (Fig. 1; see Fig. 8 for quantitative comparison). The effect of raised temperature was reversible at an intracellular pMg of 3.0 (Fig. 1A). Although a significant rundown was detected after the first temperature change, the second application of heating exhibited the same reversible response to temperature. Temperature change in the same range between 24 to 32°C produced a much more pronounced effect on the amplitude of I_Ca (Fig. 1B) when cells were dialyzed with a low-Mg²⁺ solution (pMg of 6) (Fig. 1B). In this case, intermediate temperature at 28°C could produce a gradual increase in I_Ca at an pMg of 6. Nonetheless, the effect of elevated temperature under these conditions was not reversible, i.e., returning the temperature from 32 to 24°C did not restore I_Ca to the original level at 24°C. This indicates that the enhancing effect of the low-Mg²⁺ solution (pMg = 6) on I_Ca, which has been shown previously in frog ventricular myocytes (28), can only be seen at temperatures higher than 24°C in guinea pig ventricular myocytes. In frog ventricular myocytes, the enhancing effect of low-Mg²⁺ solutions was suppressed in the presence of intracellular GTP (0.4 mM) (27). The same was true in guinea pig ventricular myocytes. Addition of 0.4 mM GTP to the low-Mg²⁺ (pMg = 6) solution suppressed the potentiation of I_Ca by low Mg²⁺ (Fig. 2). There was a statistically significant effect (P < 0.01) of GTP when I_Ca was compared (asterisk in Fig. 2; 15.3 ± 0.6 pA/pF with GTP, n = 4, and 41.2 ± 4.0 pA/pF without GTP, n = 5). Thus the potentiation of I_Ca induced by low Mg²⁺ is qualitatively the same in guinea pig ventricular myocytes at 32°C and frog ventricular myocytes at 24°C.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Influence of bath temperature and pipette free Mg2+ concentration on current through L-type calcium channels (ICa) of guinea pig ventricular myocytes. A and B: representative time course for peak ICa during step changes of bath temperature between 24 and 32°C (left). Current recordings at the indicated times (a-d) are shown along with the pulse protocol used (right). (Fast Na+ and T-type Ca2+ currents were inactivated by a 200-ms conditioning prepulse to 40 mV, and ICa through L-type Ca2+channels was elicited by a depolarizing test pulse to +10 mV; holding potential was 80 mV.) Pipette solutions contained 1 mM Mg2+ (pMg = 3; A) and 1 µM Mg2+ (pMg = 6; B).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   GTP blocks the enhancing effect of low-Mg2+ internal solution (pMg = 6) on ICa of guinea pig ventricular myocytes. A bath temperature change from 24 to 32°C (top) evoked larger ICa (bottom) in the absence of GTP () than in the presence of GTP (0.4 mM; ). Each curve gives the time course of ICa in an individual cell. The GTP effect was reproducible in several trials. In two of the trials with low Mg2+ (pMg of 6) plus GTP in the pipette solution, the temperature was changed in two steps, from 24 to 30°C and from 30 to 32°C. * Statistical significance (P < 0.01) between the groups, pMg6 and pMg6 + GTP.

Mg²+ concentration dependence of I_Ca. At 32°C, intracellular dialysis of low-Mg²⁺ solution (pMg of 5 and 6) via the whole cell patch pipette caused an increase in peak I_Ca along a biphasic time course, i.e, an early transient increase (runup) and a later sustained increase (Fig. 3A). As runup was measured at the maximum I_Ca attained earlier than 6 min for each [Mg²⁺]_i, there was a slight [Mg²⁺]_i dependency (Fig. 3C). The late phase of I_Ca increase was clearly [Mg²⁺]_i dependent. Dialysis with solutions containing relatively high [Mg²⁺]_i (pMg values of 2, 3, or 4) never induced the late phase of I_Ca increase. However, at pMg = 5, I_Ca was increased to a level comparable with that at pMg = 6 (Fig. 3A). The relationship between the Mg²⁺ concentration and the late phase of increased I_Ca is summarized in Fig. 3C. (I_Ca in the late phase was measured at 12-14 min, the length of time required for I_Ca to reach a plateau at pMg = 5 and pMg = 6.) The current-voltage relationship (current-voltage curve) of I_Ca with pMg = 6 at Fig. 3A, a and b, was compared in Fig. 3B. Depletion of Mg caused an increase in I_Ca as well as a negative shift of the current-voltage curve.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Intercellular Mg2+ dependency of peak ICa of guinea pig ventricular myocytes at 32°C. A: time course of peak ICa during internal dialysis of Mg2+ (pMg = 2-6) at a bath temperature of 32°C. Data are presented as means ± SE of ICa (n = 8, 6, 8, 7, and 8 for pMg2, pMg3, pMg4, pMg5, and pMg6, respectively). B: mean current voltage relationships with SE (n = 8) at pMg6 obtained at different times (a and b in A). C: concentration-effect curve for ICa during dialysis at the indicated intracellular [Mg2+] ([Mg2+]i). Symbols refer to the maxima of the early and late phases of increased ICa, as defined in the text. Data are presented as means ± SE of ICa from the same cells shown in A.

Phosphorylated Ca²+ channel does not respond to changes in [Mg²⁺]_i. To induce maximal phosphorylation of L-type Ca²⁺ channels, we combined extracellular perfusion of 3 µM FSK with intracellular dialyisis of the phosphatase inhibitor 10 µM okadaic acid (OA) throughout experiments. In agreement with previous findings in frog ventricular myocytes (29), this maneuver abolished the biphasic time course of I_Ca potentiation, which otherwise occurred during dialyisis of low-[Mg²⁺]_i (pMg = 5 and 6) solutions (Fig. 3A). Thus it is conceivable that depletion of Mg²⁺ led to channel dephosphorylation when OA was not included in intracellular solutions, causing a decrease in I_Ca. With the use of this protocol, we monitored peak I_Ca through presumably fully phosphorylated Ca²⁺ channels during intracellular dialysis of various [Mg²⁺]_i solutions and obtained curves of Mg²⁺ concentration dependence for the late phases of I_Ca potentiation from data at 11.2 min, as shown in Fig. 4A (open circles in Fig. 4B). The curve was not influenced by changes in [Mg²⁺]_i until [Mg²⁺]_i was raised to at least 10 mM. For comparison, the curve of Mg²⁺ concentration versus I_Ca (late phase) obtained in the absence of FSK and OA (closed squares in Fig. 4B) is replotted from Fig. 3C. This clearly indicates that phosphorylation abolished or markedly reduced the sensitivity of the Ca²⁺ channel to Mg²⁺ block, much as was previously demonstrated for the L-type Ca²⁺ channel in frog ventricular myocytes (29).

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Phosphorylation alters intracellular Mg2+ dependency of peak ICa of guinea pig ventricular myocytes at 32°C. A: time course of peak ICa during internal dialysis of Mg2+ (pMg = 2-6). The protocol was identical to that in Fig. 3A except that 10 µM okadaic acid (OA) was included in the pipette solutions and 3.0 µM forskolin (FSK) was added to the bath liquid. Data are presented as means ± SE of ICa (n = 4, 4, 3, 4, and 5 for pMg2, pMg3, pMg4, pMg5, and pMg6, respectively). B: influence of OA on the concentration-effect curve. The mean ICa at 11.2 min (; phosphorylated) have been plotted as a function of [Mg2+]i, with OA included in the pipette solution. The second curve (; dephosphorylated), which was taken from Fig. 3C (late peak), is shown for comparison.

Early transient increase in I_Ca initiated by dialysis of patch pipette solutions. After the whole cell patch-clamp configuration was established, I_Ca gradually increased (runup) and reached a peak within 5 min at 24°C (Fig. 1B and Fig. 5A). Cell rupture, which permits intracellular dialysis with pipette solutions, was a requirement for runup, because we normally did not observe this phenomenon using the nystatin-perforated patch technique (4.98 ± 1.7 pA/pF at 1.2 min vs. 5.0 ± 1.5 pA/pF at 5.4 min for frog ventricular myocytes, n = 4). Depletion of Mg may be a causative factor in runup, because Mg would be expected to diffuse into the pipette fluid down its electrochemical gradient, resulting in a reduction of Mg²⁺ block. If this were the case, peak values of I_Ca during runup should depend on the [Mg²⁺] of the pipette solution. Thus we compared peak I_Ca values as a function of the pipette [Mg²⁺] at 24°C (Fig. 5B). There was a clear dependency of the early transient increase in I_Ca on [Mg²⁺]_i at 24°C.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Intracellular Mg2+ dependency of peak ICa of guinea pig ventricular myocytes at 24°C. A: time course of peak ICa during internal dialysis of Mg2+ (pMg = 3-6) at a bath temperature of 24°C. Data are presented as means ± SE of ICa (n = 5, 3, 4, 19, and 6 for pMg, pMg3, pMg4, pMg5, pMg6, and pMg6 with GTP, respectively). GTP (0.4 mM) was included in the pipette solution for the data of pMg6 with GTP. B: concentration-effect curve for ICa (maximum of early phase vs. [Mg2+]i). Data are presented as means ± SE of ICa from the same cells shown in A. No late phase was evident in these experiments because of the low bath temperature employed. At pMg = 6, note that 0.4 mM GTP () did not produce a significant effect compared with the absence of GTP ().

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Evaluation of distinct features of early transient increase in ICa in guinea pig ventricular myocytes. A: efficiency of cAMP dialysis contained in patch pipettes into guinea pig ventricular myocytes. ICa increased as cAMP diffused into cells in a dose-dependent manner. Data are presented as means ± SE of ICa. All data were taken at 24°C using pMg3 pipette solutions but without cAMP for control (n = 4), 0.1 mM cAMP (n = 4), and 0.4 mM cAMP (n = 4) added. B: intracellular GTP effect on runup facilitated by pMg5 solution. Data are presented as means ± SE of ICa. Data were obtained with pipette solutions of pMg5 with (; n = 4) and without 0.4 mM GTP (; n = 4). There was no statistical difference between these two groups. C: runup under 100% Na+ external solution. Data were obtained with pMg3 (n = 5) and pMg5 (n = 6) solutions and are presented as means ± SE of ICa. In both conditions, runup was clearly observed and was larger with low Mg2+.

Low Mg²+ effect on frog I_Ca at reduced temperature. One of the unique properties of the guinea pig L-type Ca²⁺ channel is that the response to [Mg²⁺]_i depletion disappeared at low temperature. We determined whether L-type Ca²⁺ channels in frog ventricular myocytes also possess such a property. In the amphibian myocyte, Mg²⁺ sensitivity was present at room temperature. Therefore, we examined the effect of low Mg²⁺ (pMg = 6) on I_Ca at 17.5°C. In all fourteen trials at 17.5°C, low-Mg²⁺ solution either failed to increase I_Ca at all or induced a very small change. Averaged I_Ca at 17.5°C obtained 12.1 min after the disruption of the membrane was 6.1 ± 0.85 pA/pF (n = 14), which is clearly smaller than that obtained at 24°C (53.6 ± 4.5 pA/pF, n = 10). Estimated I_Ca at 24°C from the value at 17.5°C (6.1 pA/pF) using a Q₁₀ of pMg = 3 solution (2.05 ± 0.36, n = 4) for frog ventricular myocytes was 9.7 pA/pF, and this value was far below the one experimentally obtained with the low-Mg²⁺ (pMg = 6) solution (53.6 pA/pF). The protocol depicted in Fig. 1 was applied to frog myocytes with the exception that the temperature change was between 17.5 and 24°C; the results are shown in Fig. 7. Similar to findings in guinea pig ventricular myocytes, raising the temperature from 17.5 to 24°C produced a much larger effect at pMg = 6 (Q₁₀ = 23.0 ± 6.9, n = 6) than at pMg = 3 (Q₁₀ = 2.05 ± 0.36, n = 4) or at pMg = 6 (Q₁₀ = 3.3, n = 2) with GTP present. Moreover, the lack of reversibility of the low Mg²⁺ effect was the same as our finding in guinea pig myocytes.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Effect of temperature (Temp) changes on ICa with various internal solutions in frog ventricular myocytes. A-C: time course of changes in ICa amplitude (bottom) at the indicated bath temperature (top) and [Mg2+]i. A protocol much like the one depicted in Fig.1 was carried out. (ICa was elicited by an 80-mV depolarizing pulse of 140 ms; holding potential was 80 mV.) Current recordings were taken at the times indicated by a-f (right). Pipette pMg values were 3.0 (A) or 6.0 (B and C). In C, note the additional presence of GTP (0.4 mM) in the pipette solution. The results were similar to those obtained in guinea pig ventricular myocytes.

Temperature coefficients for various parameters of I_Ca. The Q₁₀ value for time to peak I_Ca was ~3.0 in all groups of guinea pig myocytes tested (see Fig. 8B). The effects of temperature change on inactivation kinetics were more complicated. At pMg = 6 or in the presence of FSK, I_Ca decayed along a double-exponential time course (with time constant 1_inact and 2_inact), but at pMg = 3, I_Ca decayed monoexponentially (with time constant _inact). The Q₁₀ value for _inact, which averaged 5.6 ± 0.2 ms (n = 5), was obtained from the data shown in Fig. 8C. For other conditions (pMg = 6 and FSK stimulation), faster (1_inact) and slower (2_inact) time constants and their relative amplitudes are given in the three-dimensional plot in shown Fig. 9. The relative amplitude for 1_inact was increased and that for 2_inact was decreased by the rise in temperature in both experimental conditions, whereas the temperature effect on FSK-stimulated current was much smaller. The same temperature increase diminished both 1_inact and 2_inact at pMg = 6, but this kind of effect was absent or much smaller in the case of FSK stimulation. Thus despite the complexity of analyzing the temperature dependency of I_Ca decay (particularly in experimental conditions yielding double-exponential kinetics), the results indicate that a rise in temperature accelerates the inactivation process. The relatively small temperature effect on the FSK-stimulated current is in agreement with a previous report (2).

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Quantitative comparisons of temperature effects on kinetic parameters of ICa with various internal solutions in guinea pig ventricular myocytes. A: effect of bath temperature on the amplitude of ICa. Q10 values were calculated to be 3.7 ± 0.4 (n = 4), 2.1 ± 0.03 (n = 3), 14.5 ± 0.2 (n = 6), and 3.9 ± 0.4 (n = 4) at pMg = 3, at pMg = 3 with FSK, at pMg = 6, and at pMg = 6 with GTP present, respectively. B: effect of bath temperature on the time to peak ICa. Q10 values were calculated to be 3.1 ± 0.2 (n = 4), 2.8 ± 0.4 (n = 3), 3.4 ± 0.26 (n = 6), and 3.0 ± 0.4 (n = 4) at pMg = 3, at pMg = 3 with FSK present, at pMg = 6, and at pMg = 6 with GTP present, respectively. C: influence of bath temperature on the inactivation time constant () for ICa. Decay of ICa, measured at pMg = 3, was well fitted with a single-exponential function, yielding inactivation  (left). Representative records of ICa at 24 and 32°C taken from the same cell are plotted in right, having  of 60.6 and 16.4 ms, respectively. Single-exponential curves for each time constant are superimposed on the records with solid lines. The Q10 with respect to  was calculated to be 5.6 ± 0.2 (n = 5).

View larger version (36K): [in this window] [in a new window]   Fig. 9.   Temperature effect on fast (1inact) and slow  of inactivation (2inact) in guinea pig ventricular myocytes. ICa decay was fitted with a double-exponential function at pMg = 6 (A) or at pMg = 3 with external FSK (B). The two  and their relative amplitudes were obtained at temperatures of 24 or 32°C and plotted in three-dimensional format. A: at pMg = 6, temperature rise decreased 1inact and 2inact from 46.5 ± 29.8 and 129 ± 27 ms to 10.1 ± 0.98 and 33.3 ± 5.1 ms, respectively (n = 3). B: with phosphorylation stimulated by FSK, the initial values of 1inact and 2inact were relatively small, and the effect of temperature rise (from 24 to 32°C) was also minimal, i.e., 1inact and 2inact were changed from 7.8 ± 4.2 and 67.8 ± 5.8 ms to 9.5 ± 1.8 and 37.0 ± 2.2 ms, respectively (n = 3).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We found that Mg²⁺-dependent inhibition of the L-type Ca²⁺ channel is present in both guinea pig and frog cardiomyocytes and is temperature dependent. The only difference between the frog and the guinea pig preparation was the temperature range capable of activating this Mg²⁺-dependent mechanism; i.e., the mechanism is activated above 32°C in guinea pig myocytes and above 24°C in frog myocytes. The remarkably high Q₁₀ value (14.5) for low Mg²⁺ potentiation (pMg = 6) of the I_Ca amplitude compared with the low Q₁₀ value for L-type Ca²⁺ channel kinetics may be evidence for a series of temperature-sensitive, enzymatic steps mediating the low Mg²⁺ effect.

Two kinds of Mg²+-dependent regulation. At low [Mg²⁺]_i, we observed two kinds of I_Ca facilitation: a transient early phase (runup) (taking place within <5 min from the start of intracellular dialysis) and a late phase (developing after >12 min). Both phases were Mg²⁺ sensitive but distinct from each other in that the Mg²⁺ dependency was much more manifested at low temperatures for runup and only at higher temperatures for the late phase of the I_Ca increase. Furthermore, the former mechanism was refractory to the blocking effect of GTP.

What is the mechanism for the marked temperature effect on peak I_Ca with low-Mg²+ solutions? Increasing temperature generally facilitates all kinetic processes. However, the final outcome of a temperature change in a complex system depends on the temperature sensitivity of each element in the system. For example, given similar Q₁₀ values for the activation and inactivation kinetics of voltage-gated Na⁺ channels, we would predict, using the Hodgkin-Huxley model (10), only a small increase in peak current, i.e., faster channel inactivation tends to offset faster channel activation, and a gain in peak current would be expected solely on the basis of the slight temperature dependency of the single channel conductance, which is essentially the same as that of diffusion (Q₁₀ value, ~1.5). In reality, it has been reported that the change in the open probability of L-type Ca²⁺ channels contributes to the decrease in I_Ca by the fall in temperature at a Q₁₀ of 1.5 (13, 16) in addition to a factor of unit conductance. However, the Q₁₀ value for peak I_Ca (14.5) with low-Mg²⁺ (pMg of 6) solutions, as found in the present study, is still extraordinary. This Q₁₀ value cannot be explained by a single factor such as diffusion, which lacks the necessary temperature sensitivity. Thus several factors such as enzymatic processes, lipid membrane properties, changes in channel protein conformation, or coupling between L-type Ca²⁺ channels and other proteins (e.g., sarcoplasmic reticulum Ca²⁺ channels) could be involved in this phenomenon. Allen and Mikala (3) reported on the enhanced sensitivity of human L-type Ca²⁺ channels to temperature when they were expressed in Xenopus oocytes, indicating the involvement of many factors such as membrane environment, channel assembly, and so on.

Inactivation kinetics vary with the internal environment. Generally speaking, current decay of the L-type Ca²⁺ channel in guinea pig ventricular myocytes proceeds along a double-exponential time course (2, 6, 23, 33). In this study, we found that fixing [Mg²⁺]_i at 1.0 mM (using a strong Ca²⁺ buffer of BAPTA and Mg²⁺ buffers, EDTA, and ATP) eliminated the faster time constant, leading to a monoexponential decay. This special state was perturbed by phosphorylation or reduction in [Mg²⁺]_i, i.e., either treatment created a faster component of inactivation. Do these two mechanisms (phosphorylation and Mg²⁺ depletion) ultimately cause the same effects on inactivation? There is a discrepancy of values of time constants between an pMg of 6 and 3 (FSK) at 24°C (Fig. 9). However, _inact at pMg = 6 are quite close to those at pMg = 3 (FSK) at 32°C. This is rather likely, because I_Ca of guinea pig ventricular myocytes was not stimulated with Mg²⁺-depeleting solution at 24°C as shown so far. Thus we can reasonably argue that enhanced I_Ca stimulated either with Mg²⁺ depletion or FSK show a similar _inact. These results are consistent with our hypothesis that phosphorylation modulates the channel activity by changing sensitivity to Mg²⁺ block, because phosphorylation produced a kinetic effect similar to that of low Mg²⁺.

Phosphorylation and temperature. The temperature effect on phosphorylated Ca²⁺ channels was twofold. First, channel phosphorylation is regulated by the balance between two distinct enzyme types, protein kinases and protein phosphatases. The temperature dependences of these enzymatic activities are different, so that maximal channel phosphorylation would be achieved at a temperature optimal for net phosphorylation. Second, the activating energy required to open the channel may be lower for phosphorylated channels. Thus the thermodynamic effect would be greater for unphosphorylated channels. This was the case in our study, because the Q₁₀ values for peak I_Ca were smaller in phosphorylated channels, which is consistent with Allen's results (2).

Physiological relevance. As seen in this study, temperature was an important factor in the regulation of the L-type Ca²⁺ channel in both mammalian and amphibian cardiac myocytes. Thus one must be careful to avoid ruling out mechanisms in a system without considering the influence of temperature. At the same time, the regulatory machinery of the L-type Ca²⁺ channel is not simple but is a well organized and orderly system influenced by many factors, including lipid structure, combinations of different kinds of enzymes, and so on.