INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion Conclusions References

The Ca²⁺ release channel/ryanodine receptor in sarcoplasmic reticulum (SR) plays a critical role in excitation-contraction coupling in vertebrate skeletal muscles by opening in response to sarcolemmal depolarization. Ca²⁺-induced Ca²⁺ release (CICR) is also observed with the Ca²⁺ release channel, and its properties have been extensively investigated. Cytoplasmic Ca²⁺ has a dual action on the channel activity, i.e., it is stimulating between 10⁶ to 10⁴ M and inhibitory at higher concentrations, by acting on activating and inactivating Ca²⁺ sites of the Ca²⁺ release channel, respectively (Endo, 1981; Kurebayashi and Ogawa, 1986; Meissner et al., 1986; Ogawa and Harafuji, 1990).

The properties of CICR have been investigated with several methods using different preparations, such as Ca²⁺ efflux measurement from Ca²⁺-loaded SR in skinned fiber preparations (Endo, 1981; Kurebayashi and Ogawa, 1986; Murayama et al., 1997) or isolated vesicles (Meissner et al., 1986), [³H]ryanodine binding to the Ca²⁺ release channel proteins (Ogawa and Harafuji, 1990; Murayama and Ogawa, 1996), and single-channel measurements with ryanodine receptor incorporated into planer lipid bilayers (Smith et al., 1986, 1988). Among them, skinned fiber preparations have the advantage that the organization of various biological components including SR is well maintained and that it is easy to modulate the cytoplasmic milieu by changing the composition of incubation solutions repeatedly with the same preparation. In the preparation, the channel activity has usually been determined as a rate of Ca²⁺ efflux into cytoplasmic solution from SR loaded with a specified level of Ca²⁺ according to a downhill gradient. The rate of Ca²⁺ efflux depends on both the Ca²⁺ gradient across SR membrane and permeability of the channel. It is easy to measure Ca²⁺ efflux under the condition where the Ca²⁺ gradient is large, but it would be difficult to determine the efflux rate in the presence of high concentrations of [Ca²⁺]_C where the Ca²⁺ concentration gradient is small or may be reversed. It is also impossible to determine Ca²⁺ efflux from empty SR. Kitazawa and Endo (1976) have shown that Ca²⁺ influx into empty SR with a reversed gradient was stimulated by caffeine and suppressed by Mg²⁺ or procaine, and claimed that the influx was a reversal in direction of Ca²⁺ release.

Recently modulatory effects of luminal Ca²⁺ of SR ([Ca²⁺]_L) on the Ca²⁺ release channel activity have been claimed. The reported results, however, are somewhat controversial and variable: increased Ca²⁺ release channel activity by increased [Ca²⁺]_L (Ikemoto et al., 1989; Donoso et al., 1995; Sitsapesan and Williams, 1995), decreased activity (Fill et al, 1990; Ma et al., 1988), or biphasically affected activity (Tripathy and Meissner, 1996). Various putative underlying mechanisms for those findings are also proposed: direct action of Ca²⁺ itself on ryanodine receptors from the luminal side (Sitsapesan and Williams, 1995) or from the cytosolic side accessed by Ca²⁺ coming out (Ma et al., 1988; Fill et al., 1990; Tripathy and Meissner, 1996), and indirect action via luminal proteins such as calsequestrin (Ikemoto et al., 1991; Kawasaki and Kasai, 1994).

The measurement of Ca²⁺ influx may be useful for determining the channel activity at high concentrations of [Ca²⁺]_C and/or low concentrations of luminal free Ca²⁺ to elucidate these discrepancies. In this paper we determined the properties of Ca²⁺ influx into SR in skinned fibers under a reversed Ca²⁺ gradient and showed that the Ca²⁺ influx as well as the Ca²⁺ efflux occurred mainly through the Ca²⁺ release channel. We also estimated the size of luminal Ca²⁺ binding sites in SR from the equilibrium level of calcium content (free Ca²⁺ plus bound Ca²⁺) in SR at various levels of [Ca²⁺]_C. The effect of luminal Ca²⁺ on the Ca²⁺ permeability of SR in situ was examined by comparing time courses of Ca²⁺ efflux from loaded SR and Ca²⁺ influx into empty SR and by modeling of Ca²⁺ fluxes by using the estimated parameters for the size of Ca²⁺ binding sites in SR.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion Conclusions References

Preparations

Frog (Rana japonica) was killed by decapitation, and the iliofibularis muscle was excised. A single skinned fiber (50-85 µm in diameter), isolated and mechanically split in a relaxing solution (see Table 1 for composition), was mounted in an experimental chamber at a sarcomere length of 2.6 µm (Kurebayashi and Ogawa, 1986; Murayama et al., 1997). All of the experiments were performed at a temperature of 16°C.

Solutions

Table 1 shows the composition of experimental solutions. A relaxing solution (RS) and Ca²⁺ loading solutions (L1 and L2) contained 4 mM MgATP and 1 mM free Mg²⁺, and other solutions, i.e., three kinds of washing solutions (W1-W3), test solutions (TS), and a discharging solution (DS), did not contain ATP to avoid active transport of Ca²⁺ into SR by Ca²⁺ pump. W3 and DS contained 0.1 mM EGTA and 1 µM fura 2 for detection of Ca²⁺ release from SR. A Ca²⁺ concentration in TS was achieved by mixing EGTA and CaCl₂ as described in Table 1. AMP at 4 mM was added to TS in most experiments, except where stated otherwise (cf. Fig. 5). Ionic strengths of all of the solutions were adjusted to 0.16 with KCl, and pHs were adjusted to 6.8 with KOH. Free Ca²⁺ concentration in TS was calculated using 1.048 × 10⁶ M¹ for the apparent binding constant of EGTA for Ca²⁺ (Harafuji and Ogawa, 1980). All solutions contained 2 µg ml¹ leupeptin, which kept the Ca²⁺-releasing and Ca²⁺-accumulating activities of SR in good conditions during the entire course of experiments.

Experimental setup for Ca²⁺ determination and calibration of fura 2 ratio signals

The previous experimental setup (Kurebayashi and Ogawa, 1991) was modified to enable us to measure fura 2 fluorescence (Murayama et al., 1997). Briefly, the experimental chamber (2-mm width × 30-mm length × 1-mm depth) had a bottom of clear thin glass and was fixed on a stage of an inverted epifluorescent microscope (Nikon TMD, Tokyo), which was equipped with a Spex spectrofluorometer (model CM1T11I; Spex, Edison, NJ). The skinned fiber and the surrounding solution in the chamber were illuminated with an alternating beam of 340 nm and 380 nm, and epifluorescent light longer than 420 nm within a rectangular field of 350 × 350 µm was detected by a photomultiplier. The ratio of fluorescent light of fura 2 excited at 340 nm to that at 380 nm was determined to detect Ca²⁺ concentration changes within the field. The increase in fluorescence ratio (R) that was caused by the increase in Ca²⁺ in DS was normalized by the maximum ratio difference, R_max-min, (= R_max  R_min, where R_min and R_max denote ratios on the addition of 10 mM EGTA and 10 mM CaCl₂, respectively). The ratio for a Ca²⁺-unloaded fiber in DS (R_DS) was close to R_min ((R_DS  R_min)/R_max-min  0.002).

To mimic determinations with skinned fibers, we determined a relationship between R/R_max-min and the amount of Ca²⁺ that was added to DS in the experimental chamber through a micropipette containing 10 mM CaCl₂ solution, which was connected to a pressure system Picospritzer II (General Valve Co., Fairfield, NJ). The tip (inner diameter of 1-3 µm) was placed at the left side of the monitored rectangular field in the experimental chamber. When a given period of pressure pulse was applied at a constant pressure, the ratio signal increased to reach the peak and then slowly decreased (Fig. 1 A), which was due to diffusion of Ca²⁺. The peak amplitude of R/R_max-min increased with the pulse duration. The R/R_max-min signal within 0.15 was linear to the pulse duration with an x intercept of 15 ms (Fig. 1 B). Similar results were obtained with two other micropipettes of different diameters (data not shown). The slope was steeper with a larger diameter of micropipette, but the x intercept was the same. Because a linear relation between ejected volume and pulse duration with a particular mechanical lag time is an expected performance for this pressure system as the manufacturer describes, we concluded that R/R_max-min signal up to 0.15 is linearly related to the amount of Ca²⁺ added to DS. The ejected volumes from the pipette indicated in Fig. 1 B were determined as follows. The change in the excitation spectrum (320-420 nm) of 50 µl W3 solution after ejection of 20-50 pulses followed by complete mixing was compared with that of W3 solutions containing calibrated amounts of Ca²⁺. The result indicated that a single pressure pulse of 500 ms duration ejected 600 pl of the solution through the pipette. The ejected volumes that were calculated from the varied durations of the pulse by the linear relationship were plotted on the second x axis in Fig. 1 B. Note that the lag time of 15 ms was adjusted in the second axis. In this relation, 10 pmol of CaCl₂ corresponded to R/R_max-min of 0.13. Similar relations were obtained with the other micropipettes.

View larger version (17K): [in this window] [in a new window]   FIGURE 1   Calibration of Ca2+ signal of fura 2 in our experimental setup. (A) Typical ratio signals recorded when 10 mM CaCl2 was ejected from the micropipette into DS solution, which mimics Ca2+ release from SR in skinned fiber. The ejected volume was controlled by applying pressure pulses for 100, 200, 500, or 700 ms at a constant pressure. The fluorescence ratio signal increased to the peak value and then decreased. This decrease is due to spreading out of Ca2+ from the focus, not to fluorescent bleaching. The absolute ejected volume was determined in separate experiments (see text for details). (B) Relationship of the peak level of the fura 2 ratio signal (R/Rmax-min) versus the amount of Ca2+ applied from the micropipette. Note that R/Rmax-min is proportional to the amount of applied Ca2+ within the value of 0.15.

Ca²⁺ flux measurements

Experimental protocols for determination of Ca²⁺ fluxes were similar to those described previously (Murayama et al., 1997). Before a series of experiments, a skinned fiber was treated with DS to empty the SR of Ca²⁺ and was kept in RS until use. A series of steps for Ca²⁺ influx protocol were carried out as shown in Fig. 2 A. After removal of ATP by washing in succession with solution W1 for 60 s and solution W2 for 30 s, the skinned fiber was treated with a TS containing a specified concentration of Ca²⁺ for a specified period. The fiber was then washed again with solutions W1, W2, and W3 for 30 s, 15 s, and 15 s, respectively. During the washing procedure, the amount of calcium accumulated in SR did not significantly decrease. The fiber was then challenged with DS to discharge completely releasable calcium in SR. The peak R/R_max-min in DS is indicative of the amount of calcium accumulated in SR. The treatment with DS was enough to discharge all calcium in SR, because further application of 5 µM A23187 or 1% Triton X-100 to solution W3 did not show further change in the Ca²⁺ signal (data not shown). The magnitude of R/R_max-min was at its highest (0.12) for the Ca²⁺ release from SR in DS in our experiments, which was within a linear range of the Ca²⁺-R/R_max-min relation, as indicated in Fig. 1 B.

View larger version (20K): [in this window] [in a new window]   FIGURE 2   Experimental protocols for Ca2+ influx into empty SR and for Ca2+ efflux from loaded SR. (A) Procedure for Ca2+ influx into empty SR, together with superimposed traces of Ca2+ ratio signal in DS after incubations for 0, 10, 40, and 600 s with a TS containing 3 mM Ca2+ and 4 mM AMP. (B) Procedure for Ca2+ efflux measurements with traces of ratio signal, showing the remaining Ca2+ in SR after incubation with TS for the same periods. SR was actively loaded with Ca2+ to a constant level at first, and the remaining calcium in SR was determined after TS treatment. Note that the amounts of calcium in SR achieved by either procedure reached an almost identical level after 600 s of incubation in TS.

For Ca²⁺ efflux measurements, SR in a skinned fiber was at first actively loaded to a constant level by incubation with solution L1 of pCa 6.5 for 2 min (prescriptive loading) (see Fig. 2 B). The following successive steps, including washing, treatment in TS, washing, and discharging Ca²⁺, were the same as those for influx measurement, and the amount of remaining calcium in the SR after test treatment for a specified period was determined (Fig. 2 B, traces). These influx or efflux protocols were repeated with the same fiber. In some experiments, the time courses of Ca²⁺ efflux from lower loading levels were compared to that from the prescriptive loading level. To obtain a lower loading level, the skinned fiber was incubated in solution L2 of pCa 7.0 for 60 s (for 1/3 loading) or 120 s (1/2 loading).

The prescriptive loading levels without any releasing stimulus were determined every three to five series of determinations and used as a control (prescriptive loading). In the text, the amount of calcium in SR was expressed as a value relative to the amount of the prescriptive loading level, except where stated otherwise. The prescriptive loading level amounted to 90% of the maximum accumulation obtained in solution L1. The level was between two and three times higher than the amount of calcium in the SR determined just after the skinning procedure, which nearly corresponds to the calcium content in the SR in intact fiber (data not shown).

The amount of calcium in SR at the prescriptive loading in a skinned fiber of 80-µm diameter would correspond to an impulse of ~5.3 pmol Ca²⁺ in Fig. 1, on the basis of the following findings: the monitored muscle length was 350 µm; the total calcium concentration in SR would be 30 mM (see Fig. 9); the SR volume is estimated to be 10% of the total muscle (Ogawa, 1970; Baylor et al., 1983) (0.04² *  * 0.35 * 10⁶ * 0.1 * 0.03 mol). The amount of Ca²⁺ would give a R/R_max-min of 0.065 (see Fig. 1), which is comparable to the actual R/R_max-min observed for the prescriptive loadings in our skinned fiber experiments (see Fig. 2 B).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion Conclusions References

Identification of Ca²⁺ influx pathway

Initially we examined the basic properties of passive Ca²⁺ influx into SR to confirm that the involved main pathway is the Ca²⁺ release channel. Traces in Fig. 2 A show ratio signals of fura 2 in DS after a skinned fiber with empty SR was incubated in a TS containing 3 mM cytoplasmic Ca²⁺ ([Ca²⁺]_C) and 4 mM AMP for various periods. The amplitude of the Ca²⁺ signal, which indicates the amount of calcium accumulated in SR, increased with prolongation of the incubation period. Fig. 2 B, the counterpart of Fig. 2 A, shows ordinary CICR experiments with the same preparation prescriptively loaded. We would like to point out that the remaining amounts of calcium in SR after 600 s of stimuli were similar in Ca²⁺ influx and efflux (see also Fig. 8).

Examples of time courses of Ca²⁺ influx into SR at 1 mM (open squares) and 5 mM [Ca²⁺]_C (closed circles) are plotted in Fig. 3. The amount of calcium in SR is expressed relative to the prescriptive loading level. The accumulated calcium in SR became larger with a longer incubation period and finally reached a steady level, depending on [Ca²⁺]_C. The time courses of the Ca²⁺ influxes could be fitted with an exponential equation, A * {1  exp(k_app * t)}, where A is the amount of calcium in SR at steady state and k_app is the apparent rate constant of the Ca²⁺ influx. It should be noted that no clear lag or delay was observed in the rising phase of the influx, indicating no sign of change in Ca²⁺ permeability through the SR membrane during Ca²⁺ influx into the empty SR. The steady loading level at 5 mM [Ca²⁺]_C was higher than that at 1 mM [Ca²⁺]_C, whereas k_app was ~2.5 times smaller in 5 mM [Ca²⁺]_C.

View larger version (19K): [in this window] [in a new window]   FIGURE 3   Time course of Ca2+ influx into SR. The amount of calcium in SR was determined after incubation of empty SR for various periods with a TS containing 4 mM AMP and 1 mM Ca2+ () or 5 mM Ca2+ (). Solid lines correspond to the expression Y = A * {1  exp(kapp * t)}. A and kapp for the best fit were 0.33 and 3.72 min1 for 1 mM [Ca2+]C and 0.80 and 1.43 min1 for 5 mM [Ca2+]C, respectively. The unity in "Calcium in SR" corresponds to the amount of calcium loading at pCa 6.5 for 2 min at 16°C.

The steady level of calcium in SR (A) and the apparent rate constant (k_app) are plotted as a function of [Ca²⁺]_C (Fig. 4). The steady amount of calcium in SR monotonically increased with an increase in [Ca²⁺]_C. The passive loading level at pCa 2.0 was similar to the prescriptive loading level, and a higher Ca²⁺ loading level was attained with incubation in 30 mM [Ca²⁺]_C. The apparent rate constant, k_app, on the other hand, decreased with an increase in [Ca²⁺]_C. When [Ca²⁺]_C was 30 mM, k_app was as small as 0.2 min¹, and it took more than 300 s to reach a plateau. This negative dependence of the rate constant, which corresponds to the Ca²⁺ permeability, on high concentrations of [Ca²⁺]_C was similar to the characteristics shared by CICR activities in skinned fibers, SR vesicles, and purified proteins (Endo, 1981; Meissner et al., 1986; Ogawa and Harafuji, 1990; Murayama and Ogawa, 1996; Murayama et al., 1997).

View larger version (15K): [in this window] [in a new window]   FIGURE 4   Steady amount of calcium in SR () and apparent rate constants () of Ca2+ influx into SR as a function of [Ca2+]C. The apparent rate constants were obtained by fitting time courses of Ca2+ influx into SR as shown in Fig. 3. Note that the amount of calcium in SR at the steady state increased with increase in [Ca2+]C, whereas the rate constants decreased. Each point and bar represents the mean ± SE of up to six independent determinations in different fibers.

It is well known that the Ca²⁺ release channel is activated by adenine nucleotides and inhibited by procaine or Mg²⁺. We examined the effects of these modulators on Ca²⁺ influx. The dependence of the Ca²⁺ influx rate on AMP concentration was plotted in Fig. 5. The rate of Ca²⁺ influx was very slow in the absence of AMP, although the steady level attained was similar (data not shown). The rate increased with increase in AMP concentration and reached the maximum at 4 mM or higher. The experiments described below were carried out in the presence of 4 mM AMP.

View larger version (14K): [in this window] [in a new window]   FIGURE 5   Effect of AMP on Ca2+ influx into SR. Rates of Ca2+ influx were normalized by the value in the presence of 4 mM AMP.

Fig. 6 A shows the effect of procaine on Ca²⁺ influx at various levels of [Ca²⁺]_C and compares it with the effect on Ca²⁺ release, i.e., Ca²⁺ efflux. Ca²⁺ efflux from the prescriptive loading was examined at pCa 5.3 and 3.0, and Ca²⁺ influx into empty SR was examined at pCa 3.0 and 2.0. The apparent rate constants in the presence of procaine were normalized to that in the absence of procaine at each condition. The dose dependence of Ca²⁺ influx on procaine at pCa 3.0 was very similar to that of efflux at the same pCa, and half-inhibitory concentrations were ~4 mM. Therefore the effect of procaine on Ca²⁺ fluxes was independent of the direction of flux. Furthermore, the extent of inhibition by procaine was similar regardless of [Ca²⁺]_C. Even at very high [Ca²⁺]_C, such as 10 mM, where very strong inhibition by Ca²⁺ was already obvious, procaine was as effective as at a lower [Ca²⁺]_C, at pCa 5.3, where no inactivating effect of Ca²⁺ was observed. This indicates that the inhibitory effect of procaine occurs independently of the inactivating effect of high concentrations of [Ca²⁺]_C. This result is consistent with the results of experiments with skinned fibers in that procaine suppressed bell-shaped Ca²⁺-dependent CICR activity without changing Ca²⁺ dependence (Endo, 1981) . 

View larger version (14K): [in this window] [in a new window]   FIGURE 6   Effects of procaine (A) and Mg2+ (B) on Ca2+ influx and Ca2+ efflux at various concentrations of [Ca2+]C. Time courses of Ca2+ efflux were determined at pCa 5.3 () and 3.0 (); influx was determined at pCa 3.0 () and 2.0 (). Each rate constant of fluxes was normalized to that in the absence of procaine or Mg2+.

Mg²⁺ is another well-known CICR inhibitor, and the inhibition is due to both the competition with Mg²⁺ for activating Ca²⁺ sites and the synergistic effect of Mg²⁺ on inactivating Ca²⁺ sites (Meissner et al., 1986, 1997). Consistently, Endo (1981) has reported that Mg²⁺ not only reduces CICR activity, but also shifts this Ca²⁺ dependence to higher [Ca²⁺]_C. As expected, the extent of inhibition will be dependent on [Ca²⁺]_C. In accordance with the prediction, Ca²⁺ efflux at pCa 5.3 was strongly suppressed by Mg²⁺, with a half-inhibitory concentration (IC₅₀) of less than 0.2 mM, whereas the efflux at pCa 3.0 was moderately inhibited, with an IC₅₀ of 1.5 mM (Fig. 6 B). At the same [Ca²⁺]_C of pCa 3.0, the IC₅₀ for Ca²⁺ influx into SR was 1.2 mM, which was similar to that for Ca²⁺ efflux. At pCa 2.0, the influx was clearly much less sensitive to Mg²⁺, probably indicating the same inactivation site for Ca²⁺ and Mg²⁺ in the Ca²⁺ release channel (Meissner et al., 1986). This is in contrast to procaine, which was completely independent of the inhibitory effect of [Ca²⁺]_C.

As described above, CICR modulators such as procaine and Mg²⁺ suppressed the rates of Ca²⁺ influx and Ca²⁺ efflux to a similar extent when determinations were made at the same [Ca²⁺]_C. Those inhibitors could reduce the rates of Ca²⁺ influx and Ca²⁺ efflux to less than 10% of the original activity. These results indicate that the major pathway of Ca²⁺ influx is the Ca²⁺ release channel, as is true of Ca²⁺ efflux. The measurement of the rate of Ca²⁺ influx can be a useful tool for the study of in situ activity of the Ca²⁺ release channel, especially at high [Ca²⁺]_C or at very low [Ca²⁺]_L.

Comparison of the time courses of Ca²⁺ fluxes

Ca²⁺ efflux from SR at various loading levels

1

1

1

1

View larger version (20K): [in this window] [in a new window]   FIGURE 7   Comparison of Ca2+ effluxes from SR at various loading levels. (A) Time courses of Ca2+ efflux from SR of prescriptive loading (, ) and of half the loading () were determined at pCa 5.7 using a single skinned fiber. (B) Experiments similar to those in A, except that one-third loading was carried out on a different single fiber. In both panels, data points were obtained in the order , , and .

Ca²⁺ influx into empty SR and Ca²⁺ efflux from loaded SR

1

1

View larger version (16K): [in this window] [in a new window]   FIGURE 8   Time courses of Ca2+ influx into empty SR () and Ca2+ efflux from SR at prescriptive loading () at 2 mM [Ca2+]C, which were determined in the same single fiber. The best fits (solid lines) to the experimental points were given by Y = 0.48 * {1  exp(2.05 * t)} for influx and Y = 0.52 * exp(1.99 * t) + 0.48 for efflux.

The relationship between equilibrated luminal calcium level and [Ca²⁺]_C: detection of Ca²⁺ binding sites in SR

As shown in Fig. 8, the final steady level of calcium in SR during the Ca²⁺ influx experiment was identical to that during the Ca²⁺ efflux experiment, and the level was determined by [Ca²⁺]_C, as shown in Table 2. The results shown in Fig. 4 and Table 2 are consistent with one another, suggesting the presence of Ca²⁺ binding sites, as already reported with isolated SR vesicles (Ikemoto et al., 1989; Volpe and Simon, 1991). To improve accuracy and precision in the results, similar determinations in varied [Ca²⁺]_C were carried out in the same skinned fiber, and those results with three different fibers are compiled in Fig. 9 for analysis of the Ca²⁺ binding sites. When the fiber with the prescriptively loaded SR was incubated with 10 mM [Ca²⁺]_C, the level of calcium in SR changed only slightly, whereas the final loading level became close to the prescriptive loading when the empty SR was incubated with the same solution. Therefore, the free [Ca²⁺] inside SR ([Ca²⁺]_L) at the prescriptive loading appeared to be ~10 mM. For direct comparison of amounts of calcium in SR, the equilibrated levels in SR at various pCas were normalized by the value at pCa 2.0 instead of the prescriptive loading level (Fig. 9).

View larger version (22K): [in this window] [in a new window]   FIGURE 9   Ca2+ binding sites in SR. (A) Schematic model of distribution of calcium in SR at a cytoplasmic Ca2+ ([Ca2+]C). (B and C) The relationship between equilibrated levels of calcium in SR and [Ca2+]C in skinned fibers. Calcium in SR was determined after incubation of skinned fibers in a specified concentration of [Ca2+]C for 300 or 600 s when the steady state was attained from either direction of influx or efflux. Different symbols represent different preparations. Data points are the same in B and C. Solid lines represent Y = F *{Btotal * [Ca2+]Cn/(Kdn + [Ca2+]Cn) + [Ca2+]C}, assuming that total calcium in SR is the sum of free and bound calcium. When the fitting was performed with variable n, the best fit was obtained with parameters of Btotal = 38 mM, Kd = 5.0 mM, n = 0.65 and F = 29.8 mM1, and the regression coefficient, r2, was 0.989 (B). When n was fixed at 1.0, the best fit values were Btotal = 14 mM, Kd = 1.0 mM, and F = 42.2 mM1, and r2 was 0.986 (C). Total calcium, solid lines; free Ca2+, short dashed lines; bound Ca2+, long dashed lines.

To estimate the fraction of bound to total calcium in SR, we have fit the normalized calcium content in SR (NCC) to the following equation, by assuming that total calcium in SR ([Ca]_total) is the sum of free Ca²⁺ ([Ca²⁺]_L) and bound Ca²⁺ ([CaB]_L), as shown in Fig. 9 A:

(1)

where F is a normalizing factor, and B_total, K_d, and n represent the total concentration, the dissociation constant for Ca²⁺, and a Hill coefficient of Ca²⁺ binding sites, respectively. Here we assume that [Ca²⁺]_C = [Ca²⁺]_L in the steady state, because the membrane potential across SR membrane would negligibly contribute to distribution of Ca²⁺ (Somlyo et al., 1981; Volpe and Simon, 1991). The solid curve in Fig. 9 B represents the best fitted curve, which was obtained when n = 0.65, K_d = 5 mM, and B_total = 38 mM. The Hill coefficient of 0.65 can be reasonably interpreted as the sum of multiple heterogeneous binding sites with different affinities, as discussed later in this section, although the possibility of negative cooperative binding cannot be excluded. We also made a second type of calculation by using the Hill coefficient of the unity under the assumption that the major Ca²⁺ binding sites would be homogeneous and independent in SR (Fig. 9 C). K_d and B_total obtained from the fitting were 1.0 mM and 14 mM, respectively.

In the two kinds of fittings, the square of regression coefficient, r², of 0.987 in the case of Fig. 9 C was not very different from that in the case of Fig. 9 B (r² = 0.989). It is difficult to differentiate the two alternatives from the present experimental results because of low affinities of Ca²⁺ binding sites and of unsaturating increase in free Ca²⁺. The experimental data in Fig. 9 can also be explained by a combination of intermediate values for n, K_d, and B_total, which were obtained in the experiments in Fig. 9, B and C. The relationship between total calcium in SR and [Ca²⁺]_C indicates the presence of massive Ca²⁺ binding sites in SR in skinned fibers, although it did not give a set of decisive values for binding parameters.

One of the most probable candidates of the Ca²⁺ binding sites in SR is calsequestrin (MacLennan and Wong, 1971; Ikemoto et al., 1971, 1974; Volpe and Simon, 1991). Volpe and Simon (1991) reported that the K_d, B_total, and n for calsequestrin in frog SR vesicles were 1.1 mM, 6.1 mM, and 1, respectively, in a physiological condition. On the other hand, the Ca²⁺ pump in rabbit SR is reported to have three low-affinity Ca²⁺ binding sites per molecule with a K_d of ~1 mM (Ikemoto, 1975), and its concentration in frog SR is estimated to be 0.6 mM (Ogawa, 1970). The Ca²⁺ binding sites in Fig. 9 C can be interpreted to be calsequestrin and the Ca²⁺ pump protein. Then the estimated value for B_total is 8 mM (6.1 + 0.6 * 3), which is similar to the value of 14 mM in the simulation, considering experimental errors in the estimation.

Miyamoto and Kasai (1979) described several classes of Ca²⁺ binding sites in SR. Among them, 3 sites had a K_d of 38 mM and a capacity of 144 nmol/mg protein of SR (see also Volpe and Simon, 1991). Probable Ca²⁺ binding sites may also include phospholipids such as phosphatidylserine or phosphatidylinositol, which were noticed as nonspecific binding sites (Philipson et al., 1980). These heterogeneous binding sites with various K_d values, in addition to calsequestrin and Ca²⁺-ATPase, may explain the binding profile shown in Fig. 9 B.

By using Fig. 9, we can convert a relative value of calcium in SR into total and free calcium concentration in SR. The prescriptive loading was similar to the steady state after an equilibration in pCa 2.0, when [Ca²⁺]_L = 10 mM. At the prescriptive loading, therefore, free and total calcium concentrations in SR are estimated to be ~10 mM and 24 mM (Fig. 9 C) to 33 mM (Fig. 9 B), respectively. The fraction of free to total calcium would become higher with a higher calcium loading level. For example, when the amount of total calcium in SR was decreased to one-third of the prescriptive loading, estimated free calcium and total calcium concentrations would be ~1 mM and 8-10 mM, respectively.

In intact muscle, the calcium loading level was two- to threefold lower than the prescriptive uptake level. Then the total calcium concentration in SR in intact muscle fiber is estimated to be 8-16 mM. Because the fractional volume of SR and muscle water space were estimated to be 0.11 and 0.7 of the whole fiber volume (Ogawa, 1970; Baylor et al., 1983), respectively, the value of 8-16 mM would correspond to 1.1-2.3 mmol/liter muscle water in frog skeletal muscle, which was close to the reported values in intact or cut frog muscle (1-4 mmol/liter) (Somlyo et al., 1981; Pape et al., 1995; Shirokova et al., 1996; Owen et al., 1997)

In empty SR, on the other hand, total calcium concentration should be very small, because no increase in fura 2 fluorescence signal was detected in Triton X-100 or A23187 containing solution W3. The free Ca²⁺ concentration in empty SR should be less than 20 µM, because it is the lowest Ca²⁺ concentration detectable by our experimental system.

The effects of luminal Ca²⁺ binding sites on the time course of Ca²⁺ fluxes

Ca²⁺ influx into empty SR and Ca²⁺ efflux from loaded SR

1

(2)

1

k

B  A

k

View larger version (39K): [in this window] [in a new window]   FIGURE 10   Computed time course of change in calcium in SR during Ca2+ fluxes. A and B demonstrate time courses of Ca2+ influx into empty SR and Ca2+ efflux from SR at prescriptive loading in the presence of 1 mM [Ca2+]C. Parameters for Ca2+ binding sites correspond to the results in Fig. 9, B and C, respectively: Btotal = 38 mM, Kd = 5 mM, and n = 0.65 (A) or Btotal = 14 mM, Kd = 1 mM and n = 1 (B). The intrinsic rate constant (k1) used in those simulations was 6 min1, and luminal initial free [Ca2+]L was set to zero for empty SR and 10 mM for SR at prescriptive loading. The calculated time courses of total calcium in SR were then refitted to a first-order reaction (broken lines) to obtain apparent rate constants. , , : Total, free, and bound calcium, respectively. (C) Simulated time courses of total, free, and bound Ca2+ in SR during Ca2+ efflux from the prescriptive loading (10 mM [Ca2+]L) into 0 mM [Ca2+]C. The parameters used for the simulation were Btotal = 14 mM, Kd = 1 mM, n = 1, and k1 = 6 min1. Normalized time courses of Ca2+ effluxes from SR at the prescriptive, 1/2, 1/3, and 1/6 loading levels (which correspond to the points indicated by arrows a-d) were plotted in D. The calculated values for kapp are 1.19, 0.73, 0.56, and 0.46 min1 for a (), b (), c (), and d (), respectively.

1

1

1

1

1

1

1

Ca²⁺ efflux from SR at various loading levels

1

k

A model that can explain our results

To get better agreements between our determinations and simulations, in the next model we assumed that the intrinsic rate constant k₁ is affected by [Ca²⁺]_L in the relationship, as shown in Eq. 3, but is otherwise unchanged:

(3)

where K_i is an inhibitory constant for Ca²⁺. Other parameters used were all the same as those in Fig. 10, B and C. Fig. 11 A shows a simulation for time courses of Ca²⁺ efflux from the prescriptive loading and Ca²⁺ influx into empty SR at [Ca²⁺]_C = 1 mM, corresponding to Fig. 10 B. When K_i was 10 mM or larger, the results were similar to those in Fig. 10 B. As K_i became smaller, the efflux k_app (i.e., k_app for the Ca²⁺ efflux) decreased, whereas the influx k_app did not change much. When a K_i of 2 mM was used, the efflux k_app (0.78 min¹) was very close to the influx k_app (0.80 min¹). With a K_i smaller than 2 mM, the efflux k_app became smaller than the influx k_app. Therefore 2 mM is acceptable for K_i in the case of 1 mM [Ca²⁺]_C in Table 2 (see also Fig. 11). With the successful value of 2 mM for K_i, we made similar calculations at various levels of [Ca²⁺]_C, as shown in Table 2; the results were summarized in Table 3. Within the accuracy of our determinations, this model with K_i = 2 mM satisfactorily explains the results shown in Table 2.

View larger version (26K): [in this window] [in a new window]   FIGURE 11   Inhibitory effects of luminal Ca2+ on the simulated time course of change in calcium in SR during Ca2+ fluxes. Simulation was performed by assuming that k1, the intrinsic rate constant, is negatively modulated by [Ca2+]L according to Eq. 3 (see the text for details). The effective coefficients (k'/k1) are indicated by solid lines in each panel. Parameters used for the simulations were Btotal = 14 mM, Kd = 1 mM, n = 1, k1 = 6 min1, and Ki = 2 mM. (A) Time courses of Ca2+ influx into empty SR (initial [Ca2+]L = 0 mM) and Ca2+ efflux from SR at prescriptive loading (initial [Ca2+]L = 10 mM) in the presence of 1 mM [Ca2+]C. The calculated time courses of total calcium were refitted to a first-order reaction (broken lines) to obtain apparent rate constants (see also Table 3). (B) Simulated time course of total, free, and bound calcium during Ca2+ efflux from the prescriptive loading (10 mM [Ca2+]L) into 0 mM [Ca2+]C. Note that the time course is well fitted to a first-order reaction, with a kapp of 0.5 min1 (broken line).

We also examined whether time courses of Ca²⁺ efflux from different loading levels as shown in Fig. 7 can be explained. When K_i was 2 mM, a time course of the efflux was very close to a single exponential, giving similar efflux k_app values from any Ca²⁺ loading level (Fig. 11 B). The calculated values of k_app were 0.51, 0.50, and 0.46 min¹ for the initial levels of the prescriptive, 1/2, and 1/3 loading (Table 3). When a K_i of 1 mM was used, the calculated efflux k_app from the prescriptive loading level (0.32 min¹) was smaller than that from 1/2 loading levels (0.38 min¹).

We have shown that a K_i of 2 mM gave a good fitting. We should mention, however, that a value in the range of ~2 mM for K_i may also be satisfactory. It should also be pointed out that there may be many possible ways other than this model for luminal Ca²⁺ to have an inhibitory effect. The effect of Ca²⁺ may be cooperative; the extent of the inhibition may be partial, unlike the case proposed here (100% inhibition) and others. In any event, a negative modulation of Ca²⁺ permeability by luminal Ca²⁺ is strongly suggested. It is interesting that a satisfactory value for K_i is similar to the dissociation constant for Ca²⁺ of luminal Ca²⁺ binding sites mentioned above. This suggests that some of the Ca²⁺ binding sites besides the Ca²⁺ release channel may serve the inhibitory function. The Ca²⁺ release channel would probably have two kinds of inhibitory Ca²⁺ sites, the conventional one on the cytoplasmic side and a new one on the luminal side.

	CONCLUSIONS

Top Abstract Introduction Materials & Methods Results & Discussion Conclusions References

In this article we have shown that the Ca²⁺ influx activity into empty SR reflects well Ca²⁺ release channel activity in SR in situ: it was stimulated by AMP and suppressed by procaine, Mg²⁺, and high concentrations of Ca²⁺, as was true of Ca²⁺ efflux activity in the same preparations. The effect of luminal Ca²⁺ on the Ca²⁺ release channel activity was investigated by comparing the time courses between Ca²⁺ influx into empty SR and Ca²⁺ efflux from calcium-loaded SR, and those among Ca²⁺ effluxes from SR at various loading levels. The apparent rate constant for the Ca²⁺ efflux from loaded SR was not significantly different from that for the Ca²⁺ influx into empty SR at a specified [Ca²⁺]_C. We did not find any evidence for lowered permeability of SR membrane at very low [Ca²⁺]_L compared to that at high [Ca²⁺]_L. This is different from the conclusion by Donoso et al. (1995) and Ikemoto et al. (1989), who reported lowered permeability at a low [Ca²⁺]_L. A possible explanation for this discrepancy may be the difference in preparations: they used an isolated triad or heavy SR vesicle fraction, whereas we used skinned fibers, which maintain the entire SR structure, including terminal cisternae and longitudinal tubules.

The time course of Ca²⁺ efflux from the prescriptive loading was not very different from that from one-third loading. Similar results were obtained with depolarization-induced Ca²⁺ release (DICR) from SR in intact and skinned fibers. Pape et al. (1995) reported that the relation between the rate of Ca²⁺ release by single action potential and the calcium content in SR appeared to be approximately linear in frog cut muscle fibers between 1 and 4 mmol total calcium/liter muscle water. Owen et al. (1997) reported that the depolarization of the T-system by ionic replacement could fully deplete SR and that the amplitude of depolarization-induced force response was dependent on the calcium content in SR of toad and rat peeled skeletal muscle fibers. Although CICR may be different from DICR in the mode of opening of the channel, a similar effect of luminal Ca²⁺ may be involved in the rate of Ca²⁺ release in vivo.

We have confirmed the existence of Ca²⁺ binding sites in SR from the relationship indicated in Fig. 9. Because we observed the amount of total (the sum of free and bound) calcium but not free Ca²⁺ in SR in skinned fiber preparations, we have to take the effect of Ca²⁺ buffering capacity of SR into account for analysis of the rate of Ca²⁺ fluxes. Under an assumption of no regulation by luminal Ca²⁺, the computer simulation using parameters for luminal Ca²⁺ binding sites as obtained in Fig. 9 would lead to the following conclusions: 1) the apparent rate constants (k_app) of Ca²⁺ influx into empty SR would be smaller than that of Ca²⁺ efflux from loaded SR at a [Ca²⁺]_C; 2) k_app of Ca²⁺ efflux from highly loaded SR would be larger than that from less loaded SR. These conclusions, however, are at variance with our findings. A conceivable explanation is that luminal Ca²⁺ may have a negative regulatory effect on the Ca²⁺ release channel. This is supported by simulation based on a model in which the luminal Ca²⁺ serves the inhibitory effect on the intrinsic rate constant k₁ for the Ca²⁺ release channels.

There are two possibilities for the negative effect of luminal Ca²⁺ on the Ca²⁺ release channel. It may affect the release channel from the luminal side directly or indirectly through interacting proteins. However, any evidence other than the findings presented here has not been shown for the effect of Ca²⁺ on the release channel from the luminal side. It is possible that Ca²⁺ flowing out of SR has access to cytoplasmic inactivating sites of the channel, as Tripathy and Meissner (1996) claimed, because the inactivation sites are not yet saturated in our experimental conditions. Other additional mechanisms, however, may be involved. Further investigation is required to elucidate the mechanisms of luminal regulation of Ca²⁺ release channel activity.