Relationship between Intracellular Calcium Store Depletion and Calcium Release-activated Calcium Current in a Mast Cell Line (RBL-1)*

The kinetic relationship between depletion of endoplasmic reticulum calcium stores and the activation of a calcium release-activated calcium current (I crac) was investigated in the RBL-1 mast cell line. The inositol trisphosphate receptor activator, inositol 2,4,5-trisphosphate ((2,4,5)IP3), the sarcoplasmic-endoplasmic reticulum calcium ATPase inhibitor, thapsigargin, and the calcium ionophore, ionomycin, were used to deplete stored calcium. For (2,4,5)IP3 and thapsigargin, a significant delay was observed between the initiation of calcium store depletion and the activation of I crac. However, for ionomycin, little or no delay was observed. This may indicate that a specialized subcompartment of the endoplasmic reticulum functions as a regulator of calcium entry and that this compartment is relatively resistant to depletion by (2,4,5)IP3 and thapsigargin but not to depletion by ionomycin. For all three calcium-depleting agents, the rate of development of I crac, once initiated, was relatively constant, suggesting an all-or-none mechanism. However, there were also clear experimental situations in which submaximal, graded depletion of stored calcium resulted in submaximal activation of I crac. This complex behavior could also result from the existence of a specific subcompartment of endoplasmic reticulum regulatingI crac. The kinetic behavior of this compartment may not be accurately reflected by the kinetics of calcium changes in the bulk of endoplasmic reticulum. These findings add to the growing body of evidence suggesting specialization of the endoplasmic reticulum calcium stores with regard to the control of capacitative calcium entry.

is generally accompanied by an increase in Ca 2ϩ entry across the plasma membrane. In the majority of cases, this entry seems to be signaled by depletion of the intracellular stores, a process termed capacitative calcium entry (1) or store-operated calcium entry (2). Hoth and Penner (3) first described an inward Ca 2ϩ current in RBL cells that seemed to underlie, or at least contribute to, this entry. This current they designated I crac for Calcium Release-Activated Calcium current. Although other distinguishable currents have been described that may represent capacitative calcium entry currents in other cell types (4), to date, I crac is the best characterized electrophysiological manifestation of capacitative calcium entry. Thus, its properties and modes of regulation have received considerable scrutiny by a number of laboratories. For example, Hoth and Penner (5) observed a variable latency for the activation of I crac (4 -14 s) when activated by external application of ionomycin or by break-in with IP 3 in the patch pipette. These investigators assumed that release of intracellular Ca 2ϩ by these two modes was essentially instantaneous and thus concluded that the latency observed reflected the time required for steps linking intracellular Ca 2ϩ store depletion to plasma membrane channel activation. In a more recent report, Parekh et al. (6) described an all-or-none activation of I crac by (1,4,5)IP 3 , as well as a dissociation of activation of I crac by IP 3 from the activation of Ca 2ϩ release.
In the current studies, we have further investigated the latency for activation of I crac utilizing IP 3 , the Ca 2ϩ -ATPase inhibitor, thapsigargin, and ionomycin to deplete intracellular stores and have attempted to relate these latencies to observed kinetics of intracellular Ca 2ϩ store depletion by these same reagents. Surprisingly, our findings suggest that the latency between depletion of calcium stores and I crac activation depends on the nature of the agent used to deplete the stores. We also find that the kinetics of activation are complex, with allor-none behavior in some but not in all instances. Our results may suggest the existence of a specialized, kinetically distinct subcompartment of the endoplasmic reticulum that functions as a regulator of capacitative calcium entry.
Fura-2 Loading-The cells were allowed to attach to cover slips, were mounted in a Teflon chamber, and were incubated with 3 M fura-2/AM (Molecular Probes) for 25 min at room temperature. The cells were then washed and bathed in normal external saline solution (see below) at room temperature for at least 10 min before [Ca 2ϩ ] i measurements were made.
Fluorescence Measurements-The fluorescence of the fura-2 loaded cells was monitored with a photomultiplier-based system, mounted on a Nikon Diaphot microscope equipped with a Nikon 40ϫ (1.3 N.A.) Neofluor objective. The fluorescence light source was provided by a PTI dual excitation light source equipped with a light path. The light path chopper enabled rapid interchange between two excitation wavelengths (340 and 380 nm), and a photomultiplier tube monitored the emission fluorescence at 510 nm, selected by a barrier filter (Omega). All experiments were carried out at 24°C. Calibration and calculation of [Ca 2ϩ ] i were carried out as described previously (7). * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Ruptured-patch whole-cell voltage clamp was carried out as described previously (8,9). The holding potential was 30 mV where little or no driving force for calcium entry exists. I crac was measured from the current resulting from voltage ramps between Ϫ100 to ϩ60 mV over a period of 160 ms executed every 5 s. The nonspecific current (the current before the induction of I crac or the current remaining when external Ca 2ϩ has been removed) was subtracted. All voltages were corrected for a 10-mV liquid-junction potential. Data acquisition and analysis were performed with Axopatch-1C amplifier and PCLAMP 6.1 software (Axon Instruments, Burlingame, CA). Currents were filtered at 1 kHz and digitized at 200-s intervals. Intracellular application of (2,4,5)IP 3 or external application of thapsigargin or ionomycin induced the appearance of an inward current presumed to represent I crac because (i) the current was strongly inwardly rectifying with a magnitude and current-voltage relationship similar to that previously described for I crac by Hoth and Penner (3) (not shown) and (ii) the current was seen with strong intracellular calcium buffering (10 mM BAPTA) but was lost when external calcium was removed (not shown).
Because the intracellular solutions used for Ca 2ϩ release and I crac generally differed in the major cationic species (Cs ϩ versus K ϩ ), some Ca 2ϩ release experiments were carried out utilizing the Cs ϩ -containing solutions from the I crac protocol (but with BAPTA reduced to 0.1 mM).

I crac Is Activated after an Apparent Delay when Calcium
Stores Are Emptied with (2,4,5)IP 3 -The time course of activation of calcium release and inward calcium current (I crac ) was determined in single RBL-1 cells following the introduction of (2,4,5)IP 3 into the cytoplasm via patch pipettes in the whole cell configuration. (2,4,5)IP 3 was used to minimize effects of inositol phosphate metabolism (7). The cells were held at 30 mV to minimize Ca 2ϩ entry (an approach similar to that employed by Parekh et al. (6)). After establishment of the whole cell configuration, cytosolic Ca 2ϩ rose after a short latency as a consequence of Ca 2ϩ being released from the intracellular stores by (2,4,5)IP 3 (Fig. 1A). This initial short latency is apparently the result of the time required for IP 3 to diffuse into the cell and reach a critical concentration in the vicinity of IP 3 receptors. We measured I crac activation in parallel experiments because the high levels of Ca 2ϩ buffers required in the pipette in I crac determinations (3) prevent observation of release of stored Ca 2ϩ . I crac was also activated following introduction of IP 3 after a significant latency following the establishment of whole cell configuration (Fig. 1B). However, at each of four different concentrations of (2,4,5)IP 3 the latency for I crac activation was greater than the latency for initiation of Ca 2ϩ release (Fig. 1C). We assume that the filling state of the intracellular stores is initially the same in both the Ca 2ϩ release and I crac protocols because (2,4,5)IP 3 was present in the pipette at the time of break-in, and in experiments without (2,4,5)IP 3 , I crac was not activated after prolonged dialysis, and basal [Ca 2ϩ ] i did not change appreciably. In the I crac protocol, [Ca 2ϩ ] i does not change because the intracellular solution contained ATP as well as sufficient added Ca 2ϩ to keep the free [Ca 2ϩ ] i in the physiological range (ϳ100 nM). At (2,4,5)IP 3 concentrations of 100 and 50 M, the latency for I crac activation exceeded the latency for initiation as well as for the peak of Ca 2ϩ release (Fig. 1C), suggesting a significant delay between Ca 2ϩ release and activation of I crac . Both the latency for Ca 2ϩ release and the latency for I crac activation decreased with increasing concentration of (2,4,5)IP 3 , consistent with a causal relationship between store depletion and activation of I crac (Fig. 1C). Intriguingly, the development time for I crac (from the point where I crac starts to be activated to the point where it is fully activated) remained relatively constant regardless of the concentration of The arrows indicate the intervals for estimation of latency for release and time to peak release. B, IP 3 (50 M in the pipette) activated I crac with a latency longer than that for Ca 2ϩ release. The current was measured at the potential of Ϫ100 mV from voltage ramps from Ϫ100 to ϩ60 mV and plotted versus time. The nonspecific current (the current before the induction of I crac or the current remaining when external Ca 2ϩ was replaced) was subtracted. The arrows indicate the intervals for estimation of latency of I crac and development time for I crac . C, cumulative data for the four intervals illustrated in A and B determined with four different pipette concentrations of (2,4,5)IP 3 . Each column represents average measurements from 6 to 15 cells. The latency for Ca 2ϩ release and I crac decreases with increasing concentration of IP 3 . However, a significant discrepancy exists between the time for store depletion and activation of I crac at all concentrations of IP 3 . The development time for I crac was relatively constant. IP 3 . In other words, the rate of store depletion only affects the initial latency for I crac activation, not the time course of development of I crac , at least over this (2,4,5)IP 3 concentration range. A variable delay with similar development time for I crac at different concentrations of (1,4,5)IP 3 was also reported by Parekh et al. (6).
The extent of intracellular release of Ca 2ϩ was determined in parallel experiments in which the status of intracellular stores was assessed by application of the calcium ionophore, ionomycin, 5 min after breaking into the cells (by which time I crac had reached its plateau level, Fig. 2A). Despite differing latencies for initiation of Ca 2ϩ release and I crac , at all but the lowest concentration of (2,4,5)IP 3 (10 M), the magnitude, both of intracellular release of Ca 2ϩ and of I crac , was relatively con-stant (Fig. 2, B and C). At 10 M (2,4,5)IP 3 , however, both intracellular release and steady-state I crac were clearly less than maximal (Fig. 2, B and C). In experiments with still lower concentrations of (2,4,5)IP 3 (5 M for example), a proportion of cells did not respond with release or with I crac activation. Thus, over this range of concentrations of (2,4,5)IP 3 , I crac activation seems all-or-none as described previously for (1,4,5)IP 3 (6). However, this seems to result from the fact that most concentrations of IP 3 , which are sufficient to induce release of Ca 2ϩ , induce an all-or-none release of Ca 2ϩ . At 10 M (2,4,5)IP 3 , where release is submaximal, steady-state I crac activation is also submaximal (similar finding was reported for I crac by Parekh et al. (6)). This indicates that at least over a narrow range of IP 3 concentrations, activation of I crac is likely a graded function of the extent of intracellular Ca 2ϩ store depletion (see also data below with ionomycin, Fig. 4).
A Delay for I crac Activation Is Also Seen when the Intracellular Ca 2ϩ Stores Are Depleted with Thapsigargin and Ionomycin-In the above studies, fura-2 measurement of Ca 2ϩ release is carried out under conditions of physiological low Ca 2ϩ buffering (0.1 mM BAPTA or 0.1 mM EGTA, see Experimental Procedures). However, I crac is necessarily measured under conditions of high Ca 2ϩ buffering (10 mM BAPTA) to eliminate Ca 2ϩ -activated currents and Ca 2ϩ -dependent inactivation of I crac (3). These differences in Ca 2ϩ buffering may have significant effects on the binding of (2,4,5)IP 3 to IP 3 receptors and on the amplification of Ca 2ϩ release owing to the Ca 2ϩ -induced Ca 2ϩ release (CICR) behavior of the IP 3 receptor (10). These factors could lead to a slower Ca 2ϩ release by IP 3 under the conditions for I crac measurement, thus accounting for the longer latencies we observed. To minimize these potential problems, we investigated the time course for I crac activation following depletion of stores with thapsigargin or ionomycin. Thapsigargin is a potent inhibitor of the endoplasmic reticulum Ca 2ϩ pump (11) and depletes Ca 2ϩ stores by blocking Ca 2ϩ uptake and allowing Ca 2ϩ to passively leak out. Ionomycin is a Ca 2ϩ ionophore that depletes the stores by either directly transporting ions or functioning as an ion channel. In neither case is it expected that Ca 2ϩ buffers would impede the rate of Ca 2ϩ store depletion; if anything, intracellular depletion of Ca 2ϩ would likely be augmented. We found that a delay between Ca 2ϩ release and I crac activation still seemed to be present when the store was depleted with thapsigargin ( Fig. 3) or ionomycin (Fig. 4). 2 As was seen for (2,4,5)IP 3 , the delay for activation of I crac decreased with increasing concentration of these agents. In the case of ionomycin, the development time for I crac was relatively constant for the three highest concentrations of ionomycin (58.5 Ϯ 0.9, 62.6 Ϯ 7.7, and 78.3 Ϯ 2.4 s for 500, 50, and 5 nM ionomycin, respectively), and these times are comparable with the development time when I crac was activated with (2,4,5)IP 3 . For thapsigargin, the development times for I crac were slightly longer (75.9 Ϯ 3.1, 105.2 Ϯ 12.0, 115.0 Ϯ 15.3 s for 1 M, 100 nM and 10 nM thapsigargin, respectively). Furthermore, consistent with the (2,4,5)IP 3 data, the extent of I crac activation with different concentrations of thapsigargin (Fig. 3) or ionomycin (Fig. 4) Fig. 4).

Activation of I crac Only Requires a Minimal Depletion of the Ca 2ϩ Stores, and Full Activation of I crac Does Not Require Full
Depletion of the Store-For IP 3 , thapsigargin, and ionomycin, there seems to be a significant delay between the release of intracellular Ca 2ϩ and activation of I crac . This could result either from an interval of time required to release and/or synthesize some signaling messenger or from the need to deplete intracellular stores below some critical level before the activation process begins. From inspection of the data in Figs. 3 and 4, it seems that for both agents release of Ca 2ϩ is well under way prior to the activation of I crac . However, the amount of Ca 2ϩ needed to increase [Ca 2ϩ ] i into the 100 -300 nM range is potentially very small in comparison with the total Ca 2ϩ content of intracellular stores. Thus, we attempted to determine more quantitatively the extent of depletion required to activate I crac by estimating the Ca 2ϩ store content at the time when I crac is initially activated. The latency and time course of activation of I crac for 1 M thapsigargin and 5 nM ionomycin are very similar (Fig. 4). In parallel experiments with intact RBL-1 cells, we added a high dose of ionomycin (5 M) 50 s after treatment with either 1 M thapsigargin or 5 nM ionomycin, corresponding to the time of initiation of I crac (see Fig. 3), to assess the Ca 2ϩ content of the stores and compared it with that of control cells that had not been exposed to either agent. As shown in Fig. 5, at the time when I crac was initiated by 1 M thapsigargin, significant release of Ca 2ϩ had already occurred. 100 s later, when I crac activation was maximal, a small but significant amount of stored Ca 2ϩ remained in the cells. Thus, for thapsigargin, it seems that either a significant amount of Ca 2ϩ must be released before the initiation of I crac or a significant amount of time is required for steps linking Ca 2ϩ store depletion to I crac activation.
Surprisingly, when ionomycin was used to deplete intracellular stores, a different result was obtained. 50 s after application of 5 nM ionomycin, despite the fact that a discernible elevation in [Ca 2ϩ ] i had occurred, the Ca 2ϩ content of the stores seemed to be about the same in ionomycin-treated and control cells (Fig. 6). Thus, the stores are apparently only slightly depleted at 50 s with 5 nM ionomycin. This result suggests that when ionomycin is used to deplete stored Ca 2ϩ only a very small reduction is required to initiate activation of I crac . It also indicates that for ionomycin, the delay in activating of I crac may not reflect a longer delay than that required to significantly reduce the Ca 2ϩ content of intracellular Ca 2ϩ stores.
At 150 s following addition of 5 nM ionomycin, I crac was fully activated (Fig. 4). When we used the same strategy to assess the store content at this time, we found that the Ca 2ϩ stores were depleted by about 50% (Fig. 6). This result suggests that I crac can be fully activated with only partial depletion of intracellular Ca 2ϩ stores. DISCUSSION The temporal relationship between the discharge of intracellular stores of Ca 2ϩ and the activation of capacitative calcium entry is key to understanding the nature of the signaling proc-ess. However, it is difficult to determine these two parameters under similar conditions because the very high [Ca 2ϩ ] i buffering required to detect I crac prevents detection of Ca 2ϩ release to the cytoplasm. To minimize this problem, we have relied on comparisons among three agents that cause depletion of endoplasmic reticulum Ca 2ϩ stores by clearly distinct mechanisms: IP 3 by activating a membrane receptor/ion channel; thapsigargin, which blocks the SERCA pumps that accumulate Ca 2ϩ in the endoplasmic reticulum; and ionomycin, which passively transports Ca 2ϩ down its concentration gradient. Although we expect experimental conditions such as Ca 2ϩ buffering to affect release of Ca 2ϩ by IP 3 , we expect this to be much less of a factor with thapsigargin and ionomycin. Nonetheless, we cannot know with absolute certainty that this is so. Ideally, one would like to be able to assess the Ca 2ϩ content of intracellular stores in the experiments in which I crac is measured. Recently described technologies may permit such a determination in the near future (12).
In experiments utilizing (2,4,5)IP 3 as an activator of intracellular IP 3 receptors, we observed a clear distinction between FIG. 5. Status of intracellular Ca 2؉ stores 50 and 150 s after addition of 1 M thapsigargin. The status of intracellular stores was assessed by addition of 5 M ionomycin 50 s (above) after addition of 1 M thapsigargin (dashed line), at which time I crac is just beginning, and 150 s (below), at which time I crac activation is maximal (Fig. 3). Comparison of the responses to ionomycin in control cells (no addition of thapsigargin) indicates that at 50 s, substantial depletion of stores has occurred, and by 150 s, near maximal depletion has occurred. Each trace represents an average of data obtained from 5 to 10 cells analyzed in separate experiments.
FIG. 6. Status of intracellular Ca 2؉ stores 50 and 150 s after addition of 5 nM ionomycin. The status of intracellular stores was assessed by addition of 5 M ionomycin 50 s (above) after addition of 5 nM ionomycin (dashed line), at which time I crac is just beginning, and 150 s (below), at which time I crac activation is maximal (Fig. 4). Comparison of the responses to high ionomycin in control cells (no addition of 5 nM ionomycin) indicates that at 50 s, very little depletion of stores has occurred, and at 150 s, only about 50% of maximal depletion has occurred. Each trace represents an average of data obtained from 5 to 10 cells analyzed in separate experiments. the time required for detectable release of stored Ca 2ϩ and that required for activation of I crac . That is, significantly shorter intervals were required for mobilization of intracellular Ca 2ϩ stores than for activation of I crac . As discussed above, because different Ca 2ϩ buffering conditions were necessarily used for these two determinations, it is not clear whether the latency for release of Ca 2ϩ was identical in the two experimental conditions. In fact, one might expect augmentation of the Ca 2ϩ release process with minimal Ca 2ϩ buffering through calciuminduced calcium release. However, if that were the case, one might also expect that concentrations of (2,4,5)IP 3 inducing only partial depletion of stores with high concentrations of intracellular buffers might cause complete depletion with lower buffer concentrations. From the data in Fig. 2, this does not seem to be the case; 25-100 M (2,4,5)IP 3 induced maximal responses in the high and low buffer conditions (I crac in the former, depletion in the latter), whereas 10 M (2,4,5)IP 3 induced a partial activation with both buffer conditions.
Differences in Ca 2ϩ buffering should be less of an issue for experiments utilizing thapsigargin or ionomycin. These agents deplete Ca 2ϩ stores by distinct and passive mechanisms, such that cytoplasmic Ca 2ϩ buffering should affect the kinetics of depletion minimally if at all. From inspection of the time courses of I crac activation and of [Ca 2ϩ ] i increase in Figs. 3 and 4, it seems that Ca 2ϩ release from the endoplasmic reticulum does indeed precede I crac by some tens of seconds. However, the data in Fig. 6 show that for low concentrations of ionomycin at least, the level of cytoplasmic Ca 2ϩ can be a poor indicator of the extent of depletion of Ca 2ϩ stores. Despite a significant rise in cytoplasmic Ca 2ϩ 50 s after addition of 5 nM ionomycin, the total stored Ca 2ϩ content of the endoplasmic reticulum was changed minimally. These results indicate that especially with low concentrations of Ca 2ϩ -depleting agents, changes in [Ca 2ϩ ] i can misrepresent the extent of changes in intracellular Ca 2ϩ stores. Thus, for this concentration of ionomycin, and in contrast to the findings with (2,4,5)IP 3 and thapsigargin, there may be very little delay between the fall in Ca 2ϩ content of the endoplasmic reticulum and the initiation of I crac . In an earlier report, McDonald et al. (13) reported minimal latency for I crac activation when (1,4,5)IP 3 was rapidly released within Jurkat T cells by flash photolysis.
What then is the meaning of the delay between Ca 2ϩ store depletion and I crac activation seen with (2,4,5)IP 3 and thapsigargin? As discussed above, for (2,4,5)IP 3 , the difference may reflect differences in rates of Ca 2ϩ discharge with the two different intracellular Ca 2ϩ buffering systems used. But perhaps a more fundamental difference among these three modes of Ca 2ϩ release is that only ionomycin can be assumed to release Ca 2ϩ in a spatially nonspecific manner throughout the cell. In other words, because of its presumed mechanism of action, we expect ionomycin to release Ca 2ϩ from all components or regions of the endoplasmic reticulum with similar facility. However, for (2,4,5)IP 3 and thapsigargin, this may not be the case. (2,4,5)IP 3 will cause activation of IP 3 receptor channels and cause discharge of Ca 2ϩ only at the specific sites where these receptors are located. Thapsigargin will lead to depletion of stores by passive leak of Ca 2ϩ following inhibition of SERCA pumps. Virtually nothing is known about the channels mediating this presumably IP 3 -insensitive or basal movement of Ca 2ϩ , but such sites could also be localized in a nonhomogeneous manner in specific regions of the endoplasmic reticulum. Thus, the current findings may indicate that a specific subfraction of the endoplasmic reticulum regulates the Ca 2ϩ channels underlying I crac and that as Ca 2ϩ is depleted from the endoplasmic reticulum through IP 3 receptors, or through the leak pathway involved with thapsigargin action, this specific subcompartment is more slowly depleted than the majority of the endoplasmic reticulum. To our knowledge, this is the first evidence for differential effects of thapsigargin in subcompartments of the endoplasmic reticulum. The suggestion of a specific subfraction of endoplasmic reticulum involved in the regulation of I crac is consistent with the finding that for both ionomycin and thapsigargin, full depletion of the endoplasmic reticulum store is not required for maximal activation of I crac . In at least one earlier report, specialization of the endoplasmic reticulum Ca 2ϩ stores with respect to regulation of capacitative Ca 2ϩ entry has been suggested (14). It was suggested that the subfraction of endoplasmic reticulum coupled to capacitative calcium entry was at most 30% of the total thapsigargin-sensitive Ca 2ϩ stores.
The current findings indicate that the kinetics of activation of I crac are complex. Over certain concentration ranges with all three modes of activation, the time course of I crac activation is relatively constant. This may simply result from the fact that in most experimental situations, release of Ca 2ϩ occurs more rapidly than the steps involved in signaling I crac . This may also be suggestive of an all-or-none mechanism of activation, as proposed by Parekh et al. (6). However, graded activation of I crac was observed with 10 M (2,4,5)IP 3 , as well as in an study earlier utilizing cyclopiazonic acid to deplete endoplasmic reticulum calcium (14). As in the current study, Parekh et al. (6) concluded that a small component of the total intracellular Ca 2ϩ stores regulates I crac , but this conclusion was based on findings that low concentrations of (1,4,5)IP 3 seemed to activate Ca 2ϩ release without activating Ca 2ϩ entry. In our study, we have exclusively utilized (2,4,5)IP 3 , a poorly metabolizable analog of (1,4,5)IP 3 , to avoid possible complications of differential metabolism under different experimental conditions. With (2,4,5)IP 3 as the mobilizing ligand, no dissociation between the concentrations required for release and those required for I crac activation were observed (Fig. 2). However, it is the differential latencies observed for ionomycin, thapsigargin, and (2,4,5)IP 3 that lead us to propose a specialized, quantitatively minor component of the endoplasmic reticulum as the site of control of capacitative calcium entry and I crac . If in fact a small compartment of the endoplasmic reticulum is responsible for regulation of I crac , it is possible that the kinetics of I crac activation reflect the kinetic behavior of this small pool of Ca 2ϩ and that these kinetics are not accurately reflected by the average time course of changes in cytoplasmic and stored Ca 2ϩ in the cell. Future work must concentrate on experimental dissection of these functionally distinguishable subcompartments of endoplasmic reticulum Ca 2ϩ stores if we are to fully understand how the plasma membrane Ca 2ϩ movements underlying I crac and capacitative calcium entry are regulated.