Introduction

Top Summary Introduction Materials & Methods Results Discussion References

The ability of ryanodine receptors to mediate multiple partial releases of Ca²⁺ from internal stores in response to step increases in agonist concentrations has been documented by several groups (1-3). This phenomenon bears similarity to the "quantal" release of Ca²⁺ from internal stores sensitive to InsP₃ (4). The mechanism or mechanisms responsible for these partial releases of Ca²⁺ remain highly controversial, with two main theories propounded. It has been proposed that receptors of heterogeneous agonist sensitivity are grouped together in isolated stores such that some stores release all their Ca²⁺ at low agonist concentration, and other stores follow suit only at higher agonist concentrations. In contrast to such static receptor sensitivity, it has been proposed that some other feedback mechanism regulates the agonist sensitivity of even individual release channel receptors. In its original form, this proposal involved luminal Ca²⁺ being the feedback system (5). Experimental demonstrations of such dynamic receptor sensitivity included the observations of a low level of steady state flux of Ca²⁺ from InsP₃-sensitive stores (6) or the dynamic adjustment of Ca²⁺ sensitivity of individual ryanodine receptors in planar lipid bilayers (7). The latter result was obtained under conditions in which luminal Ca²⁺ concentration was low and would not have varied.

In view of the postulated involvement of luminal Ca²⁺ in such dynamic sensitivity and the known sensitivity of both InsP₃ and ryanodine receptors to cytoplasmic Ca²⁺ concentration (8), we investigated further the role of Ca²⁺ in the quantal release of Ca²⁺ from internal stores involving ryanodine receptors. Our earlier study (3) showed that multiple releases take place regardless of whether released Ca²⁺ is resequestered, suggesting that Ca²⁺ redistributions do not control the partial release behavior. The results of the experiments reported here show that Ca²⁺ concentration changes on either side of the ryanodine receptor can modulate incremental Ca²⁺ release behavior, in particular by altering agonist sensitivity, but that these effects by themselves are unable to account for the stopping and restarting of release. Results with inhibitors provide additional strong evidence against multiple partial releases being due purely to true quantal behavior and in favor of a more dynamic contribution to incremental release behavior, one that does not rely primarily on luminal Ca²⁺ levels.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Heavy microsomes and purified terminal cisternae were prepared from rabbit skeletal muscle as described in Dettbarn et al. (3). Protein concentrations were determined using Coomassie Protein Assay Reagent (Pierce, Rockford, IL) using bovine serum albumin as a standard.

Spectrophotometric measurements of SR Ca²⁺ uptake and release were performed in a diode array spectrophotometer (HP 8451A; Hewlett Packard, Palo Alto, CA). The standard assay medium included 100 mM KCl, 20 mM K-MOPS, 2.5 mM potassium phosphate, 1 mM Mg-ATP, 5 mM Na₂ phosphocreatine, 20 µg/ml creatine phosphokinase, and 0.2 mM antipyrylazo III, pH 6.8. Some experiments (see Figs. 1, 2, 3, 4) used no phosphate, with pH 6.95 medium at 37°. Extravesicular Ca²⁺ concentration changes were followed by measuring antipyrylazo III absorbance at A₇₁₀-A₇₉₀.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Caffeine sensitivity of SR stores is affected by the luminal Ca2+ content. A, Multiple Ca2+ releases in response to step increases in caffeine concentration. Rabbit microsomes (600 µg) were preloaded with 6.25 nmol of Ca2+ in 1.25-nmol increments as described in the text. Subsequently, the vesicles were challenged with increasing concentrations of caffeine as indicated. Caffeine was never removed from the cuvette. B and C, Multiple Ca2+ releases in response to caffeine when the vesicles were loaded with more Ca2+ (12.5 nmol in B, 25 nmol in C). , 1.25-nmol Ca2+ additions. Arrows, 0.6 mM caffeine additions.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   The decrease in luminal Ca2+ caused by a submaximal caffeine-induced Ca2+ release cannot account for release termination. A, Heavy microsomes (380 µg) were preloaded with eight 12.5-nmol increments of Ca2+ (100 nmol total) and then challenged with 5 mM caffeine, releasing ~35 nmol Ca2+. B-D, Heavy microsomes (380 µg) were preloaded with only 62.5 nmol (B), 37.5 nmol (C), or 12.5 nmol (D) of Ca2+ and then challenged with 5 mM caffeine. Release still took place. , 12.5-nmol Ca2+ additions. Arrows, 5 mM caffeine additions. Traces are representative of experiments performed in triplicate under each condition.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Homogeneously loaded SR vesicles still display multiple partial releases of Ca2+. Terminal cisternae at ~15 mg/ml were equilibrated overnight with 5 mM Ca2+/45Ca-containing solution as detailed in the text. After 100-fold dilution into EGTA-containing solution (), in some cases the vesicles were challenged first with 8 mM caffeine (); in some of these experiments, the caffeine concentration was raised to 22 mM () 40 sec later. All experiments were performed at least in triplicate. Error bars, mean ± standard deviation. All data points at the same time point are statistically different from one another as assessed with a Student's paired t test (p < 0.05). Curves were drawn by eye.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Elevation of resting cytoplasmic Ca2+ levels increases agonist sensitivity. A, Heavy microsomes (380 µg) preloaded with 50 nmol of Ca2+ were then treated with 12.5 nmol of Ca2+ just before challenge with 1 mM caffeine. B, Microsomes were treated in the same way except that the caffeine was added before the last Ca2+ addition. , 12.5 nmol Ca2+ additions. Arrows, 1 mM caffeine additions. Traces are representative of experiments performed at least in triplicate, including (B) experiments (not shown) in which 62.5 nmol of Ca2+ was preloaded before caffeine addition to compensate for any additional Ca2+ taken up (in A) before caffeine addition. Regardless of whether 50 or 62.5 nmol of Ca2+ had been preloaded, caffeine still failed to elicit release unless the extravesicular Ca2+ was elevated.

Certain ⁴⁵Ca efflux determinations were carried out in the same medium at the same temperature but with ⁴⁵Ca during the initial phase of preloading of endogenous Ca²⁺ associated with the sample. The additions of 1-3 mM Ca²⁺ were then made, followed by the addition of 5 mM EGTA, appropriately alkalinized to maintain pH of the final mixture. Efflux was monitored by measuring the Ca²⁺ remaining in the vesicles by first quenching the reaction through the addition of 100-µl aliquots to 0.9 ml of ice-cold 100 mM KCl, 20 mM MOPS, 1 mM EGTA, 10 mM MgCl₂, and 20 mg/liter ruthenium red, pH 6.8. At the conclusion of each run, the quenched 1-ml mixtures were filtered by conventional manifold filtration with Millipore HAWP filters (0.45-µm effective pore size) and washing with three 3-ml washes of ice-cold quench medium.

Other ⁴⁵Ca efflux determinations involved passive loading of purified terminal cisternae with ⁴⁵Ca. This was achieved by first sedimenting terminal cisternae (15 mg/ml) out of their storage medium containing 10% sucrose with a 5-min full-speed spin in an airfuge (Beckman Instruments, Palo Alto, CA); resuspension in sucrose-free 100 mM KCl and 20 mM MOPS, pH 6.8, with airfuging again to wash; and resuspension in its original volume of 100 mM KCl and 20 mM MOPS, pH 6.8. To this suspension, 5 mM Ca²⁺ spiked with ⁴⁵Ca was added, and the mixture was equilibrated overnight on ice. For efflux determinations, 10 µl of the mixture was diluted into 1 ml of 100 mM KCl, 20 mM MOPS, and 20 mM EGTA, pH 6.9 at room temperature. Three 100-µl aliquots were withdrawn at 10-sec intervals, and each was added to 0.9 ml of quench medium as described above, followed in some cases by the addition of 2-8 mM caffeine from a concentrated stock at t = 40 sec. Three additional 100-µl aliquots were then withdrawn and quenched at 10-sec intervals, followed in some cases by elevation of the caffeine concentration to 22 mM and withdrawal of three additional aliquots for quenching and subsequent filtration as described above. All isotope data were normalized with respect to the initial (t = 10 sec) time point in each series.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Partial releases of SR Ca²⁺ in response to step changes of caffeine concentration are shown in Fig. 1. Because we prefer to reserve the term quantal to refer to one specific and as-yet-unproved explanation for the phenomenon, we will refer to such releases interchangeably as partial, incremental, or quantal throughout this text. In Fig. 1, multiple partial releases were obtained under conditions in which released Ca²⁺ was given time to be resequestered. Note that the first three additions of caffeine failed to generate any detectable release of Ca²⁺, and a large release with a plateau was obtained only on the fifth caffeine addition. If the amount of Ca²⁺ preloaded into the vesicles is increased (Fig. 1, B and C), not only are the releases of Ca²⁺ larger in amplitude, but the caffeine concentration required to initially activate release is reduced. Thus, the threshold concentration of agonist (caffeine) required to elicit release is lowered by increased luminal Ca²⁺ content. Three additional sets of experiments were carried out with 6.25, 12.5, and 18.75 nmol of Ca²⁺ preloading. In all experiments with 6.25 nmol of Ca²⁺ preloading, a large release with a plateau was obtained only on the fifth caffeine addition, whereas large releases were always obtained with the third caffeine addition if 12.5 nmol of Ca²⁺ had been preloaded and on the second caffeine addition when 18.75 nmol of Ca²⁺ had been preloaded. These differences are highly significant as assessed using a Student's paired t test.

The results of the experiments in Fig. 1 were qualitatively consistent with the original Irvine hypothesis of luminal Ca²⁺ controlling quantal release, but they certainly did not prove the hypothesis. To more quantitatively assess the possible contribution that this change in sensitivity might make to subsequent incremental release, we determined how much of the Ca²⁺ taken up had been released by caffeine. If the Irvine hypothesis were to explain quantal release, then preloading the vesicles so that their luminal content would be the same as that experienced at the time of termination of release should result in a failure of the same caffeine concentration to elicit release.

This test was applied in Fig. 2. First heavy microsomes were loaded with 100 nmol of Ca²⁺ and challenged with 5 mM caffeine (Fig. 2A). This concentration of caffeine was chosen to give submaximal release (the release seen at 3 mM was just detectable). The response was a release of ~35 nmol of Ca²⁺. We then preloaded the vesicles with progressively less Ca²⁺ (62.5, 37.5, or 12.5 nmol of Ca²⁺) before challenging them with the same concentration of caffeine. As seen in Fig. 2, a release of Ca²⁺ was observed under all these conditions, contrary to the predictions of the Irvine hypothesis. Similar results were obtained with another sample that released less Ca²⁺ in response to 4 mM caffeine (not shown) and with purified terminal cisternae that released a greater proportion of the Ca²⁺ taken up (not shown). Because vesicles with even less Ca²⁺ in them than after a release remained responsive to the same submaximal concentration of caffeine, any decrease in caffeine sensitivity caused by a release did not seem to be sufficient by itself to account for the termination of release.

A separate test was made to determine whether heterogeneously loaded stores could account for multiple partial releases of Ca²⁺ in response to incremental caffeine additions. To perform this test, purified terminal cisternae were passively equilibrated in 5 mM Ca²⁺/⁴⁵Ca-containing solutions to ensure that all vesicles were uniformly loaded with Ca²⁺. Then, the vesicles were challenged with two steps of increasing concentration of caffeine. Although the threshold concentration of caffeine required to activate release may differ in the absence of ATP, the experiment diagrammed in Fig. 3 demonstrates that multiple responses are obtained. Similar results were obtained using 2 or 5 mM caffeine for the first caffeine challenge (not shown). Thus, the multiple responses to caffeine cannot be accounted for by populations of vesicles that are differentially loaded with Ca²⁺.

However, because the effect of luminal Ca²⁺ was at least qualitatively consistent with the Irvine hypothesis, it remained possible that some other feedback system acting in concert with the luminal Ca²⁺ effect could result in release termination. The most obvious candidate to explore for such a contribution is the extravesicular (cytoplasmic) Ca²⁺ concentration because it increases whenever a release is taking place.

Accordingly, we preloaded vesicles to the same extent with Ca²⁺ and then exposed them to Ca²⁺. Finally, the vesicles were challenged with step increases in caffeine concentration. If Ca²⁺ was administered in a fashion that it was not able to be appreciably sequestered before caffeine addition, then the threshold caffeine concentration required to activate release was lowered. Vesicles unresponsive to 1 mM caffeine at ambient resting Ca²⁺ concentration became responsive if Ca²⁺ was elevated just before caffeine addition (Fig. 4). This result was not surprising in that it probably represents a corollary to the much earlier observation that caffeine enhances Ca²⁺ sensitivity (9). In fact, it is clear from Fig. 4A that caffeine administered before Ca²⁺ addition led to a larger absorbance increase from the Ca²⁺ addition, as if some Ca²⁺-induced Ca²⁺ release had indeed occurred. This result was not in agreement with a contributory role for elevated bulk cytoplasmic Ca²⁺ in terminating release. However, high local Ca²⁺ concentrations around the mouth of a conducting release channel might be more prone to induce inactivation of that channel because such local Ca²⁺ concentrations could easily exceed those of the Ca²⁺ concentrations administered in Fig. 4.

If the vesicles were exposed to only a 2-4-fold higher extravesicular Ca²⁺ concentration, another effect was encountered. Releases evoked by even high caffeine concentrations were slowed and attenuated (Fig. 5). These were Ca²⁺ concentrations only slightly higher than those achieved at the peak of most releases (15-20 µM total Ca²⁺, which we calculate to be equivalent to 5-7 µM free Ca²⁺ under our experimental conditions at pH 6.8). The downward-going artifacts observed on caffeine additions in the presence of elevated Ca²⁺ might have partially obscured some remaining release.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Further elevation of resting cytoplasmic Ca2+ levels decreases subsequent caffeine-induced Ca2+ release. Heavy microsomes (380 µg) preloaded with 50 nmol of Ca2+ were then treated with 0, 25, or 50 nmol of Ca2+ just before the addition of 10 mM caffeine. Downward shifts in traces on caffeine addition are due to caffeine reduction of antipyrylazo III sensitivity to Ca2+. , 12.5-nmol Ca2+ additions. Arrows, 10 mM caffeine additions. Due to its interaction with antipyrylazo III, caffeine induces a downward deflection in traces that is larger in the presence of high Ca2+. Traces are representative of at least three determinations under each condition.

To rule out further release by caffeine under such circumstances and to determine whether this inhibitory effect was due to CICR depletion of the vesicles or to an inactivating effect of the elevated extravesicular Ca²⁺, equivalent experiments were performed with purified terminal cisternae using the Ca²⁺ ionophore A23187 to assess the Ca²⁺ remaining in the SR after the different treatments. As seen in Fig. 6, A and B, caffeine application reduced the A23187-releasable fraction considerably, but so did before Ca²⁺ addition (Fig. 6C). Much less Ca²⁺ was left to be released by A23187 subsequent to the caffeine (4.10 ± 0.31 nmol) or Ca²⁺ addition in Fig. 6C (4.52 ± 0.67 nmol versus control 33.69 ± 0.05 nmol), and thus substantial CICR had occurred in response to the Ca²⁺ addition, probably leading to depletion rather than inactivation.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   CICR takes place even when vesicles are challenged with high cytoplasmic Ca2+. Purified terminal cisternae (300 µg) added at the first upward deflection were allowed to take up endogenous Ca2+ associated with the sample and assay solution and then challenged with 10 mM caffeine (A), 2 µM A23187 (B), or 50 nmol of Ca2+ (C). At the end of each trace, 2 µM A23187 was added, followed by 12.5 nmol of Ca2+. , 12.5 nmol Ca2+ additions. Arrow, 10 mM caffeine additions. Arrowheads, 2 µM A23187 additions. Experiments shown are representative of at least three separate determinations under each condition.

In principle, high local Ca²⁺ concentration, which only occurs while a release channel is still conducting, could, together with a decrease in luminal Ca²⁺, lead to release termination. When the release terminates, the high local Ca²⁺ concentration would dissipate very rapidly, and the system might then soon be susceptible to reactivation easily because the decrease in luminal Ca²⁺ alone had not been sufficient to prevent reactivation. To determine whether any Ca²⁺-induced inactivation of release had also occurred, we determined whether high Ca²⁺ administered exogenously would lead to inactivation before depletion. As seen in Fig. 7, Ca²⁺ elevations of even 3 mM failed to terminate release of ⁴⁵Ca before completion, and subsequent addition of EGTA (to reduce the cytoplasmic Ca²⁺ to a level where no inactivation would take place) failed to reactivate release because little Ca²⁺ remained in the SR. Similar results were obtained with 1 mM Ca²⁺ (not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Effect of high Ca2+ and its subsequent reduction on efflux of 45Ca. Terminal cisternae (300 µg) were actively preloaded with endogenous Ca2+ spiked with 45Ca in the presence of ATP. Aliquots were removed for Millipore filtration as described in the text. After three base-line readings, terminal cisternae treated as above were exposed to 3 mM total Ca2+ (2 mM free Ca2+) at t = 40 sec. Subsequent addition of EGTA sufficient to lower the free Ca2+ concentration to 1.5 µM at t = 80 sec failed to reactivate release. The experiment was performed four times with results expressed as mean ± standard deviation. Curve was drawn by eye.

The results thus far suggest that neither luminal nor cytoplasmic Ca²⁺ changes can account for multiple partial releases. Because Irvine's luminal Ca²⁺ regulation hypothesis represented the original alternative to true quantal behavior, does our dismissal of the luminal Ca²⁺ explanation represent indirect evidence in favor of true quantal behavior? To test this premise directly, we performed an additional set of experiments.

Microsomes were loaded with Ca²⁺ and then challenged with caffeine or caffeine in the presence of concentrations of tetracaine that only partially inhibited release. Two different concentrations of caffeine were used: one (4 mM) that elicited a clearly submaximal rate of release, and one (10 mM) that exhibited substantially larger, more prolonged release with a faster rate of release. A peculiarity of the tetracaine inhibition of release seen in Fig. 8 was that the dose-response curve (not shown) for tetracaine was unusually steep and clearly depended on the caffeine concentration. True quantal behavior would predict that the same vesicles that released (completely) in the presence of caffeine alone would also release (completely) in the presence of the low affinity antagonist tetracaine and that the amount released should be the same in the presence of tetracaine, even if the rate of release was reduced. This behavior was not observed experimentally. Instead, the amount of Ca²⁺ released seemed to be decreased as much as the rate of release (Fig. 8). These results suggest that some time-dependent process (either inactivation or adaptation) must have caused the release to terminate. Furthermore, it is unlikely that release termination was dependent on the length of time that conducting release channels were exposed to high local Ca²⁺ during release. Had this been the case, a 50% inhibition in release rate would have dictated a doubling of time necessary for any time- and Ca²⁺-dependent process to take place, whereas lower release rates (4 versus 10 mM caffeine or tetracaine versus controls) were associated with more rapid cessation of release.

View larger version (14K): [in this window] [in a new window]   Fig. 8.   Effect of tetracaine on caffeine-induced Ca2+ release. Heavy microsomes (380 µg) were preloaded with 100 nmol of Ca2+ in eight 12.5-nmol increments and challenged with 4 (left) or 10 (right) mM caffeine. Top, absence of tetracaine. Experiments were then repeated with 10, 15, and 20 µM tetracaine (left; concentrations as indicated) or 20, 40, and 60 µM tetracaine (right; concentrations as indicated). Traces in the presence of tetracaine demonstrated not only reduced rates of release of calcium but also reduced amounts of Ca2+ release. The control release with 4 mM caffeine represented 10 nmol of Ca2+; that elicited by 10 mM caffeine was 13.5 nmol of Ca2+. The tetracaine-induced decreases in release amplitudes shown are representative of at least three determinations of each set of conditions. With different samples, the concentration of tetracaine needed to achieve these effects differed slightly, but tetracaine always reduced release amplitudes, and more tetracaine was required to inhibit release at higher caffeine concentrations.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

The role of luminal Ca²⁺. Our result that termination of release occurs under Ca²⁺ loading conditions where the same stimulus can still elicit release (Fig. 2) leads us to conclude that luminal Ca²⁺ by itself does not control quantal Ca²⁺ release of RyRs, as in the original Irvine InsP₃ receptor hypothesis. However, as suggested in that hypothesis, a decrease in luminal Ca²⁺ does provide a negative feedback to modulate release by increasing the threshold concentration of agonist required to elicit release. We used the concept of altered threshold to describe our results because they do not strictly discriminate between a change in affinity (sensitivity) of the RyR from a decreased cooperativity of binding. For the sake of convenience in the Discussion, we use the term "sensitivity" in its most general sense, to include either possibility. Our results are consistent with the positive feedback of increased intraluminal Ca²⁺ noted with isolated SR vesicles (10). Results obtained with ryanodine receptors studied in planar lipid bilayers with high intraluminal Ca²⁺ have been more varied. In the absence of ATP, high intraluminal Ca²⁺ has been reported to decrease conductance (11) or to have no effect (12), but in the more physiological presence of ATP, high intraluminal Ca²⁺ has a stimulatory effect on P_o (12), which would be in accord with our vesicular results, also obtained in the presence of ATP. Finally, the caffeine sensitivity of Fura-2-loaded adrenal chromaffin cells was enhanced by greater luminal Ca²⁺ content (1). Thus, a decrease in luminal Ca²⁺ is not the controlling factor that terminates release for ryanodine receptors.

The role of cytoplasmic Ca²⁺ elevations. Previously, we noted that multiple releases were obtained from SR regardless of whether extravesicular Ca²⁺ was allowed to remain high and concluded that it was unlikely that extravesicular (cytoplasmic) Ca²⁺ plays a significant role in controlling release (3). This conclusion still holds for bulk cytoplasmic Ca²⁺ levels. The experiments presented here suggest that even local cytoplasmic Ca²⁺ concentrations in the vicinity of a conducting release channel play no greater role. Although elevated local Ca²⁺ in the vicinity of a reconstituted single channel can inactivate it (11, 17, 18), our results suggest that any inactivation taking place in our vesicle studies is too incomplete to terminate release before vesicle depletion (Fig. 7). Our results with tetracaine and different caffeine concentrations also argue against negative feedback by released Ca²⁺: if release did inactivate or adapt before completion, the inactivation would have to have been independent of Ca²⁺ flux through the channel because the reduction in flux by tetracaine or lower caffeine did not obviously slow inactivation. Thus, local Ca²⁺ concentration changes by themselves also cannot account for multiple partial releases. As a consequence, our results do not support application to the RyR of the model of Swillens et al. (19), which suggests that Ca²⁺ conduction through a release channel is the critical factor in terminating release.

Adaptation, true quantal behavior, or both?. True quantal release in which discrete stores with different static sensitivities individually release all of their Ca²⁺ or none at all would not account by itself for our results with tetracaine because in that case the individual stores that were sensitive to caffeine should have continued to release all their contents, albeit at a reduced rate. Furthermore, we provided previously evidence with isolated SR that caffeine can release some Ca²⁺ and that when the caffeine is diluted and its concentration subsequently restored, further release is obtained (3), another argument against true quantal release. Both arguments would apply to any static sensitivity explanation for incremental release, whether it involved different sensitivities due to differences in luminal load (current study), heterotetrameric release channels (30, 31), or different release channel densities (32).