J Biol Chem, Vol. 274, Issue 47, 33327-33333, November 19, 1999

Regulation of Ryanodine Receptor Opening by Lumenal Ca²⁺ Underlies Quantal Ca²⁺ Release in PC12 Cells^*

Schuichi Koizumi^§, Peter Lipp^¶, Michael J. Berridge, and Martin D. Bootman^**

From the  Laboratory of Molecular Signalling, The Babraham Institute, Babraham Hall, Cambridge, CB2 4AT United Kingdom, the  Department of Zoology, University of Cambridge, Downing Street, Cambridge, CB2 3EJ United Kingdom, and the ^§ Division of Pharmacology, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya, Tokyo 158, Japan

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Graded or "quantal" Ca²⁺ release from intracellular stores has been observed in various cell types following activation of either ryanodine receptors (RyR) or inositol 1,4,5-trisphosphate receptors (InsP₃R). The mechanism causing the release of Ca²⁺ stores in direct proportion to the strength of stimulation is unresolved. We investigated the properties of quantal Ca²⁺ release evoked by activation of RyR in PC12 cells, and in particular whether the sensitivity of RyR to the agonist caffeine was altered by lumenal Ca²⁺. Quantal Ca²⁺ release was observed in cells stimulated with 1 to 40 mM caffeine, a range of caffeine concentrations giving a >10-fold change in lumenal Ca²⁺ content. The Ca²⁺ load of the caffeine-sensitive stores was modulated by allowing them to refill for varying times after complete discharge with maximal caffeine, or by depolarizing the cells with K⁺ to enhance their normal steady-state loading. The threshold for RyR activation was sensitized ~10-fold as the Ca²⁺ load increased from a minimal to a maximal loading. In addition, the fraction of Ca²⁺ released by low caffeine concentrations increased. Our data suggest that RyR are sensitive to lumenal Ca²⁺ over the full range of Ca²⁺ loads that can be achieved in an intact PC12 cell, and that changes in RyR sensitivity may be responsible for the termination of Ca²⁺ release underlying the quantal effect.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

A change in cytosolic Ca²⁺ concentration (Ca²⁺_cyt)¹ serves as a signal for modulating a wide range of cellular activities (1-3). A major mechanism for increasing Ca²⁺_cyt is release of Ca²⁺ from internal stores (endoplasmic or sarcoplasmic reticulum; ER or SR) via inositol 1,4,5-trisphosphate receptors (InsP₃R) and ryanodine receptors (RyR). (1, 4, 5). These Ca²⁺ release channels are widely expressed, and have structural and functional homology (1, 4, 5). Activation of RyR or InsP₃R can give rise to spatially and temporally complex Ca²⁺ signals such as Ca²⁺ waves and oscillations (2, 6, 7).

Several studies have revealed that RyR and InsP₃R release Ca²⁺ from intracellular stores in direct proportion to the strength of stimulation (8-20). Such graded or "quantal" Ca²⁺ release is paradoxical since both InsP₃R and RyR display the autocatalytic property of Ca²⁺-induced Ca²⁺ release, which could be predicted to release the entirety of the intracellular Ca²⁺ stores once release is initially activated.

Although, as originally envisaged (8), the term "quantal Ca²⁺ release" has been used to imply the all-or-none release of discrete Ca²⁺ pools, many workers have used the term to simply denote situations where the release of Ca²⁺ from intracellular stores occurs in a graded manner. Other nomenclature, such as "incremental Ca²⁺ release" (10) and "partial Ca²⁺ release" (21) have been used to describe graded Ca²⁺ release and avoid mechanistic connotations, but these terms all describe the same basic phenomenon.

The mechanism underlying quantal Ca²⁺ release is unclear, although several schemes have been proposed, including inactivation/adaptation of InsP₃R and RyR (22-25), all-or-none release from functionally discrete Ca²⁺ stores bearing InsP₃R and RyR with distinct sensitivities (8, 17, 26), control of InsP₃R and RyR opening by the Ca²⁺ concentration within the lumen of the ER/SR (19, 27-29), and a compensatory increase in Ca²⁺-ATPase activity (30).

Although all of the proposed schemes have received some experimental support, none of them has as yet been universally accepted as the mechanism underlying quantal Ca²⁺ release (for review, see Refs. 21, 31, and 32). In the present study, we investigated the properties of quantal Ca²⁺ release evoked by activation of RyRs in intact PC12 cells, and in particular whether the sensitivity of RyRs to the agonist caffeine was altered by the lumenal Ca²⁺ load. Our data suggest that changing the loading status of the Ca²⁺ stores from almost empty to maximally full can substantially enhance the sensitivity of RyRs to the agonist caffeine. The positive correlation between Ca²⁺ release and the ER Ca²⁺ content supports the idea that decreases in lumenal Ca²⁺ underlies quantal Ca²⁺ release.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- Cells were cultured as described previously (33) with minor modification. PC12 cells were obtained from the National Institute of Health Sciences (Tokyo, Japan). All cells used in the present study were between passages 53 and 68. Cells were grown on tissue culture flasks (75 cm²) in 85% RPMI growth medium (Life Technologies, Inc.) supplemented with L-glutamine, 10% horse serum, 5% fetal calf serum, 100 units/ml penicillin, 100 µg/ml streptomycin and buffered with 2.2 g/liters NaHCO₃. Cells were incubated in a humidified atmosphere (95% air and 5% CO₂) at 37 °C, and the medium was changed 3 times per week. Cells were plated onto poly-L-lysine-coated glass coverslips at a density of 1 × 10⁶ cells/dish 2 days before use.

Measurement of Ca²⁺_cyt in Single Cells-- Cells were incubated with 3 µM fura-2AM for 45 min at room temperature, and then washed with Krebs-Ringer buffer (KR buffer) of the following composition (in mM): NaCl 145, KCl 5, CaCl₂ 3, MgSO₄ 1, NaH₂PO₄ 1.2, glucose 10, and Hepes 20, pH 7.3. Cells were further incubated with KR buffer for another 30 min to allow hydrolysis of fura-2. For the Ca²⁺_o-free experiments, we used a medium where Ca²⁺ was omitted and 2 mM EGTA was added (Ca²⁺_o-free KR buffer). For the solution containing a high concentration of KCl, 70 mM NaCl of the KR buffer was replaced with an equal concentration of KCl (75 mM final K⁺ concentration).

A coverslip bearing the adherent cells was mounted on the stage of a Nikon diaphot, inverted epifluorescence microscope. Fluorescent images were obtained by alternate excitation at 340 and 380 nm (40 ms each wavelength) using twin xenon arc lamps (Spex Industries Inc., Edison NJ). The emission signal at 510 nm was collected by a charge-coupled device intensifying camera (Photonics Science, Robertsbridge, United Kingdom), and the digitized signals were stored and processed using an Imagine image processing system (Synoptics Ltd., Cambridge, UK) as described previously (13). Stimuli and reagents were added to the cells either as a bolus by pipette, or using solenoid-switched gravity-fed perfusion tubes positioned near to the cells. All experiments were performed at room temperature (22 ± 2 °C).

Calculation of Actual Ca²⁺ Release-- One of the aims of the present study was to investigate the effect of the ER Ca²⁺ load on the sensitivity of Ca²⁺ release, and to see whether Ca²⁺ release terminated after the lumenal Ca²⁺ concentration fell. Since differences in the amplitudes of caffeine-evoked Ca²⁺ signals are not necessarily a reliable measure of relative Ca²⁺ store contents, we estimated the actual amount of Ca²⁺ release according to the following equation:

(Eq. 1)

where Ca²⁺_cyt is the maximal Ca²⁺_cyt rise above the prestimulated level, v is cell volume calculated by measuring the diameter of individual cells and assuming a spherical shape, and  is the endogenous Ca²⁺ binding capacity, i.e. bound Ca²⁺ over free Ca²⁺. The value of  was taken as 75 (34). This method assumes that Ca²⁺ release terminates at the peak of the response, and that the Ca²⁺ release is much faster than the removal of Ca²⁺_cyt.

Differentiation of PC12 Cells-- Differentiation of the PC12 cells was induced by reducing each serum concentration by 50%, and by adding 100 ng/ml nerve growth factor (7 S fraction, TOYOBO Co. Ltd., Japan). The cells were maintained for at least 7 days in the nerve growth factor-containing medium before use. The medium was changed every 2 days.

Materials-- Caffeine and ryanodine were from Sigma, fura-2 AM was from Molecular Probes (Eugene, OR), and cell culture materials were from Life Science Technologies, Inc. Where possible, analytical quality reagents were used.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Quantal Ca²⁺ Release from RyR in PC12 Cells-- Application of increasing caffeine concentrations (1-40 mM) to fura-2-loaded PC12 cells evoked a dose-dependent Ca²⁺_cyt increase (Fig. 1A), with a half-maximal caffeine concentration of ~11 mM, and maximal release at 40 mM caffeine. A similar graded response to caffeine was observed in the presence or absence of extracellular Ca²⁺ (Ca²⁺_o) (Fig. 1B).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Concentration-response relationship for caffeine-evoked Ca2+ release from RyRs in PC12 cells. Panel A illustrates the typical response of fura-2-loaded PC12 cells to increasing caffeine concentrations (filled bars) in Ca2+o-free medium (hatched bars). Ca2+ was readmitted to the medium between caffeine stimulations to allow the stores to refill. The trace represents averaged responses from 30 cells. Panel B shows concentration-response curves for caffeine-evoked Ca2+ signals, illustrating that the majority of the caffeine-stimulated Ca2+ signal arises from Ca2+ release. In separate experiments no difference was seen in the Ca2+cyt increases evoked by 40 or 80 mM caffeine, indicating that 40 mM caffeine was maximal (data not shown). Curves were fitted using a Hill equation. The half-maximally effective caffeine concentrations was ~11 mM in the presence or absence of Ca2+o (n = 18).

To determine whether or not such graded release was quantal, the cells were stimulated continuously with various caffeine concentrations until the Ca²⁺_cyt responses had declined. The cells were subsequently tested with a maximal caffeine dose. In keeping with the quantal concept, the low caffeine concentrations were incapable of depleting the entire caffeine-sensitive Ca²⁺ pool despite the prolonged application in Ca²⁺_o-free medium (Fig. 2A). The Ca²⁺_cyt signal evoked by low caffeine concentrations was inversely proportional to that evoked by the maximal caffeine concentration (Fig. 2B).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Low caffeine concentrations release only a fraction of the caffeine-sensitive intracellular Ca2+ pool. Panels A and B depict the reciprocal relationship between low caffeine concentrations and a maximal caffeine concentration, in terms of depleting the caffeine-sensitive Ca2+ pool. The concentrations of caffeine used for the initial stimulation are shown on the left of the individual curves in A. Each trace shows the average of 24-36 cells. For B, Ca2+ release was calculated as the relative Ca2+cyt increase over basal (data are mean ± S.E.; n = 24-36). Open circles show the percentage response stimulated by the initial addition of caffeine, and closed circles show the relative response evoked by a subsequent application of 40 mM caffeine. Caffeine was applied in Ca2+o-free medium. Panel C illustrates the typical variability in threshold for caffeine-evoked Ca2+ release in single PC12 cells.

The graded recruitment of the caffeine-sensitive Ca²⁺ pool in PC12 cells could also be observed by successive application of increasing caffeine concentrations in Ca²⁺_o-free medium (Fig. 2C). The threshold concentration for activating Ca²⁺ release varied between 1 and 10 mM caffeine. However, after reaching the threshold, every cell tested (n > 200) responded to subsequent caffeine applications with a partial release of the Ca²⁺ pool, and required 40 mM caffeine for total depletion. Cells that had been pretreated with 40 mM caffeine in Ca²⁺_o-free medium gave a lesser response to a subsequent application of 1 µM thapsigargin (data not shown). This indicates that the Ca²⁺ released by caffeine was not resequestered by the Ca²⁺ stores, but was exported from the cells.

Adapatation of RyRs to Ca²⁺_cyt Does Not Account for Quantal Ca²⁺ Release-- Although the caffeine-evoked steady-state activity of RyR in bilayers has been recorded for many tens of minutes (35), these channels have been shown to display a much higher open probability upon initial activation by Ca²⁺_cyt, after which the P_o rapidly declines if the stimulus is still present (36). Since caffeine activates RyR by sensitizing them to Ca²⁺_cyt (37), it is plausible that such adaptation could occur in the presence of caffeine, and could therefore account for the quantal responses depicted in Figs. 1 and 2. Indeed, adaptation is probably the most commonly suggested mechanism for quantal Ca²⁺ release from RyRs (22, 23, 38, 39).

To examine the contribution of adaptation to the quantal responses observed from PC12 cells, we stimulated cells with caffeine in a pulsatile manner. Repeated application of 40 mM caffeine at 3-min intervals (30 s caffeine applications) evoked consistent Ca²⁺_cyt increases (Fig. 3A). Shortening the time between caffeine applications caused the responses to decrease in amplitude, most likely due to the inability of the Ca²⁺ pool to fully reload in less than 3 min (see below). Extending the interval between caffeine applications to greater than 3 min did not significantly increase the Ca²⁺_cyt response (Fig. 3A). These data indicate that the RyRs in PC12 cells are able to maximally respond providing there is an interval of at least 3 min between successive stimulations. Similar repetitive responses were obtained using lower caffeine concentrations (data not shown). Therefore, if the RyRs did adapt to the caffeine stimulus, such adaptation was completely reversed in the 3-min interval before the next caffeine application.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Pulsatile caffeine application evokes quantal Ca2+ release. Panel A depicts the typical response of a single PC12 cell to repetitive 40 mM caffeine applications. Shortening of the interval between caffeine applications reduced the amplitude of subsequent responses, while lengthening the stimulus interval up to 3 min gave reproducible maximal responses. Panel B illustrates that the entire caffeine-sensitive Ca2+ store can be discharged by repetitive applications of 40 mM caffeine in a Ca2+o-free medium, and that lengthening of the interval between the stimulus does not recover the response. Panel C shows that pulsatile application of a single PC12 cell with 5 mM caffeine in a Ca2+o-free medium up to the point where the response does not to release all of the caffeine-sensitive Ca2+ pool.

Pulsatile application of 40 mM caffeine in Ca²⁺_o-free medium evoked a series of responses with progressively diminishing amplitudes (Fig. 3B), consistent with the observation that the Ca²⁺ released by caffeine is exported from the cell. In addition, these data indicated that negligible refilling of the intracellular Ca²⁺ stores took place in the Ca²⁺_o-free medium. Stimulation of the cells by pulsatile application of 5 mM caffeine in Ca²⁺_o-free medium (30-s application; 3-min interval) also evoked a series of responses with diminishing amplitudes. However, this was not due to complete depletion of the caffeine-sensitive Ca²⁺ pool, as in the case of a maximal caffeine concentration, since subsequent application of 40 mM caffeine evoked a further robust response. If the RyR adapted or were inactivated by the presence of caffeine, it would be expected that each 5 mM caffeine stimulation would have progressively released the intracellular Ca²⁺ pool until it was fully depleted. In contrast, our data suggest that adaptation to a cytosolic stimulus does not underlie quantal Ca²⁺ release from RyR in PC12 cells.

Regulation of RyR Sensitivity by lumenal Ca²⁺ Content-- The data presented above indicates that the caffeine-sensitive Ca²⁺ pool can be released in a concentration-dependent manner (Fig. 1). Each concentration of caffeine releases a fraction of the store, and then fails to have any effect, but leaves a smaller Ca²⁺ pool for the next caffeine concentration to discharge (Fig. 2). Since Ca²⁺ within the lumen of the stores has been proposed to regulate the activation of RyR (40-42), we therefore investigated the changes in lumenal Ca²⁺ content caused by various caffeine concentrations, to see whether termination of Ca²⁺ release was correlated with lumenal Ca²⁺ load.

Caffeine concentrations of 1, 5, 10, or 20 mM were applied to cells in a pulsatile manner until there were no further responses (Figs. 4, A-D). This point was taken to be the threshold lumenal Ca²⁺ concentration at which the RyR were no longer sensitive to the stimulating caffeine concentration. The total caffeine-releasable store was then emptied by application of 40 mM caffeine. The threshold lumenal Ca²⁺ content was estimated by converting the peak response to 40 mM caffeine into a value for "actual Ca²⁺ release" (see "Experimental Procedures"). The averaged responses (Fig. 4E) show that the quantal Ca²⁺ release pattern evoked by 1-20 mM caffeine is associated with a concentration-dependent reduction in the lumenal Ca²⁺ load over a ~10-fold range. These data are consistent with the idea that Ca²⁺ release evoked by low caffeine concentrations becomes inhibited as the lumenal Ca²⁺ concentration decays.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Decrease in lumenal Ca2+ during caffeine stimulation. In the experiments depicted in panels A-D, cells were stimulated with pulsatile applications of caffeine (as shown by the solid bars) until the responses stopped, and then by 40 mM caffeine to assess the remaining Ca2+ load of the intracellular pool. E, the amplitude of the Ca2+cyt response to 40 mM caffeine was converted into actual Ca2+ release as indicated under "Experimental Procedures." The Ca2+ released by 40 mM caffeine represents the fraction of the Ca2+ pool that was not mobilized by the previous applications of lower caffeine concentrations.

To further investigate the regulation of RyR sensitivity by lumenal Ca²⁺, we modulated the Ca²⁺ load within the stores. This was done by treating cells with 40 mM caffeine in Ca²⁺_o-free medium to completely empty their stores, and then subsequently allowing them to refill for variable periods of time. Reloading of the caffeine-sensitive Ca²⁺ pool to a steady state level required ~300 s, with a half-time of ~60 s (data not shown). In addition, the steady-state loading of the stores could be enhanced by approximately 40% following a brief depolarisation of the cells with high a K⁺ solution (see "Experimental Procedures") (data not shown).

The effect of different store loads on the responsiveness to caffeine was examined as shown in Fig. 5. Prior to reloading, the caffeine-sensitive Ca²⁺ stores were depleted by sequentially applying increasing caffeine concentrations in Ca²⁺_o-free medium. Following reloading, the same caffeine solutions were applied and the Ca²⁺_cyt signals at each caffeine concentration were monitored. The example traces in Fig. 5A indicate the typical pattern of responsiveness observed. In the absence of store loading, there was no effect of the second sequence of caffeine solutions (Fig. 5A, a). Responsiveness to caffeine returned when the cells were allowed to reload for variable times (Fig. 5A, b-e). The most effective reloading was achieve by stimulating with KCl to depolarize the membrane (Fig. 5A, f). Enhanced Ca²⁺ store loading had two significant effects; the threshold for caffeine-evoked Ca²⁺ release declined (Fig. 5B), and lower caffeine concentrations released greater proportions of the stored Ca²⁺ (Fig. 5C). The former effect indicates that increased lumenal Ca²⁺ load increases the sensitivity of the RyR to caffeine. The typical threshold caffeine concentration was 10 mM after a 10-s reloading period, while cells commonly responded to 1 mM caffeine after KCl treatment (Fig. 5B). The apparent lack of responsiveness of cells with low lumenal Ca²⁺ loads to the lesser caffeine concentrations was not due to our inability to monitor the small changes in Ca²⁺_cyt that would arise from poorly loaded cells. For example, with a 10-s reloading period, responses to 20 and 40 mM caffeine were easily apparent (Fig. 5A, b), and since 5 mM caffeine, for example, releases on average ~40% of the total caffeine-sensitive Ca²⁺ pool (Fig. 2B) we would have been able to monitor a release triggered by 5 mM caffeine above the noise (S.E. typically ± 0.2 nM) in our system.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Sensitization of caffeine responsiveness by increased luminal Ca2+ loading. The traces in A, a-f, illustrate typical responses observed in single PC12 cells after reloading of the caffeine-sensitive Ca2+ pool to different extents. At the start of the experiment, the caffeine-sensitive Ca2+ pool was discharge by sequential application of increasing caffeine concentrations in Ca2+o-free medium (as shown by the filled bars). After a brief period of recovery, the Ca2+o was readmitted to the extracellular medium for varying times (a, no reloading; b, 10 s; c, 30 s; d, 60 s; e, 300 s; f, 300 s plus KCl). The cells were then rechallenged with caffeine in Ca2+o-free medium. Panels B and C illustrate the effect of varying the reloading period on the responsiveness of the cells (B) and the fractional release of Ca2+ at each caffeine concentration (C). The data show mean ± S.E. of 45-112 cells for each reloading time.

The second effect of enhanced lumenal Ca²⁺ loading was that the fractional release of Ca²⁺ shifted to lower caffeine concentrations (Fig. 5C). With 5 mM caffeine, for example, the fractional release of Ca²⁺ increased ~5-fold from 12.7% after a 10-s refilling to 65.1% following KCl treatment. The effects of KCl on the loading of the stores and the enhanced sensitivity of cells to caffeine were both transient. Postincubation of cells in control buffer for 5 min following KCl treatment reversed both the enhanced loading and the increased sensitivity (data not shown).

Even at steady state conditions there was considerable variability in the actual Ca²⁺ content of the stores between individual cells. Such variation could underlie the heterogeneity of the responses of naive cells to caffeine, such as the different thresholds for caffeine-induced Ca²⁺ release (Fig. 2C). We therefore investigated whether the normal variability of the lumenal Ca²⁺ load affected the sensitivity of the cells to caffeine. Cells in Ca²⁺_o-free medium were sequentially stimulated with 1, 5, 10, 20, and 40 mM caffeine, using the protocol depicted in Fig. 2C. The Ca²⁺ release evoked by each caffeine application was summed to give a measurement of the starting Ca²⁺ content of the stores in each cell, and to calculate the fraction of Ca²⁺ release evoked by each caffeine concentration. Consistent with the effects of changing the lumenal Ca²⁺ by varying the store reloading period (Fig. 5), there was a positive correlation between the endogenous lumenal Ca²⁺ content and fractional Ca²⁺ release with low caffeine concentrations (Fig. 6). The cells with the highest starting Ca²⁺ content gave the highest fractional Ca²⁺ release in response to low caffeine concentrations. These data indicate that within the normally variable Ca²⁺ load of the PC12 cells, those cells possessing the highest lumenal Ca²⁺ content will be the most sensitive to caffeine.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Differences in steady-state loading underlie the variable sensitivity of naive PC12 cells to caffeine. PC12 cells were treated with increasing caffeine concentrations, using the same protocol as shown in Fig. 2C. The individual responses to 1, 5, 10, 20, and 40 mM caffeine were converted into actual Ca2+ release, as described under "Experimental Procedures," and summed to give a measure of the total content of the caffeine-sensitive Ca2+ pool in each cell (denoted as total Ca2+ content on the abscissa). The fractional Ca2+ release evoked by 1, 5, 10, and 20 mM caffeine were then calculated. The curves were fitted using a Hill equation. The inset shows the individual data points for the fractional Ca2+ release evoked by 5 mM caffeine.

Lumenal Ca²⁺ Depletion by Tunneling of Ca²⁺ in the ER Can Desensitize Unactivated RyR-- Since Ca²⁺ can apparently readily diffuse or "tunnel" within the lumen of the ER (43), Ca²⁺ release in one region of a cell should lead to a drop in lumenal Ca²⁺ concentration throughout the cell. Such a decrease in lumenal Ca²⁺ would be predicted to reduce the sensitivity of all RyR, irrespective of their position relative to the initial Ca²⁺ release.

In order to evoke Ca²⁺ release in only part of a PC12 cell, we differentiated the cells for 1 week using nerve growth factor to stimulate neurite outgrowth. There were no detectable differences in Ca²⁺ signaling between the undifferentiated and the differentiated PC12 cells; both displayed quantal Ca²⁺ release patterns over the same range of caffeine concentrations (data not shown). Using a microperfusion pipette situated within a fast bulk flow, proximal neuritic regions were repeatedly stimulated with 5 mM caffeine in Ca²⁺_o-free medium (Fig. 7A). As expected, the pulsatile caffeine application caused depletion of the Ca²⁺ stores within the stimulated region (Fig. 7B, b). However, subsequent bath application of the same caffeine concentration to the whole cell was unable to evoke Ca²⁺_cyt increases in any part of the cell (Fig. 7B, c). The lack of Ca²⁺_cyt increase following bath application of 5 mM caffeine was not due to complete emptying of the intracellular stores, since all regions of the cell were able to respond to subsequent application of 40 mM caffeine. As the entire cell was sensitive to 5 mM caffeine under steady-state conditions (cf. first caffeine application in Fig. 7B, b and c), these data suggest that the release of Ca²⁺ from the region of the cell stimulated with pulsatile 5 mM caffeine applications was able to lower the sensitivity of RyRs throughout the entire cell.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Tunneling of Ca2+ within the ER lumen of PC12 cells can desensitize RyR in unstimulated cell regions. The cell outline in panel A depicts the orientation of the cell body and proximal neurites relative to the microperfusion system (marked as "pipette") and direction of "bulk flow." Panel B illustrates the stimulus regime (B,a), and the Ca2+cyt responses from the proximal neurite region (B,b) and cell body (B,c) (shown by the regions bounded with dashed lines in A). A brief bath stimulation with 5 mM caffeine in Ca2+o-free medium using the bulk flow elicited a Ca2+cyt increase in both regions of the cell. Partial reloading of the stores by readmission of Ca2+o for 120 s allowed further responses in the proximal neuritic region upon focal application of 5 mM caffeine (B,b). Subsequent bulk flow stimulation with 5 mM caffeine did not evoke a Ca2+cyt increase in either region of the cell (B, b and c), yet a subsequent application of 40 mM caffeine was able to evoke a substantial global response.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Quantal Ca²⁺ release is an outstanding problem in the Ca²⁺ signaling field in that it is an almost universal observation made using many different cell types for over a decade, yet there is little consensus concerning the mechanism underlying the graded release of Ca²⁺ from intracellular stores (31, 32). Both InsP₃R and RyR have been found to display quantal Ca²⁺ release, and although the majority of studies have investigated InsP₃R, it seems that both Ca²⁺ channels can give analogous quantal release patterns. These observations are perhaps not surprising given the conserved structural and functional homology between InsP₃R and RyR (see Introduction), and they point to a common mechanism that can terminate Ca²⁺ release from the ER or SR.

The inability of low levels of stimulation to release all of the intracellular Ca²⁺ pool could conceivably have a trivial explanation, such as a compensating increase in Ca²⁺-ATPase activity (30) or Ca²⁺ pool fragmentation (44). However, many of the observed quantal responses cannot be accounted for by these simple possibilities. Instead it appears that the quantal effect is most likely due to abrupt changes in the open probability of Ca²⁺ release channels, either caused by changes in lumenal Ca²⁺ concentration (19, 12, 27), or following adaptation/inactivation of InsP₃R and RyR (22, 23, 25). Alternatively, quantal Ca²⁺ release has been proposed to be due to progressive recruitment of functionally discrete Ca²⁺ stores that have different intrinsic sensitivities, and which release their Ca²⁺ in an all-or-none manner (8, 13, 14, 17, 26, 45). In the present study, we investigated the mechanisms underlying quantal responses from RyRs in caffeine-stimulated PC12 cells, and in particular the roles of adaptation and lumenal Ca²⁺ in limiting Ca²⁺ release.

Adaptation is a novel form of channel inactivation whereby the activity of Ca²⁺ release channels declines during prolonged exposure to activator stimuli. This mechanism was first proposed to account for the graded release of Ca²⁺ from the SR in cardiac myocytes during different depolarization steps (36), and was subsequently suggested to also underlie quantal responses from RyR in other cell types (23, 39, 42). Intrinsic inactivation of InsP₃R has also been reported, and proposed as a quantal mechanism (Refs. 24 and 25; but see Refs. 45-49).

Adaptation or inactivation of Ca²⁺ release could plausibly account for quantal Ca²⁺ release during prolonged stimulation. However, it is important to note that adaptation or inactivation of both InsP₃R and RyR to cytosolic stimuli reverses rapidly upon removal of the stimulus. Since caffeine evoked consistent responses when applied in a pulsatile manner under conditions where the Ca²⁺ stores could refill (Fig. 3A), it appears that adaptation or inactivation cannot have a long-lasting effect on the RyR in PC12 cells. It is therefore difficult to reconcile the lack of response to pulsatile applications of low caffeine concentrations in Ca²⁺_o-free medium (Fig. 3C) with a scheme involving adaptation to caffeine. Similar results have been reported for RyR in intact caffeine-stimulated chromaffin cells (17), and for InsP₃R in intact (18) or permeabilized (19, 50) cells following pulsatile application of agonist or InsP₃, respectively.

Although the experiments using pulsatile application of stimuli are rather simple, they do argue that adaptation or inactivation of RyR to a cytosolic stimulus is not essential for quantal Ca²⁺ release. Our previous demonstration of such responses during pulsatile application of caffeine to adrenal chromaffin cells was discounted on the basis that it applied only to the RyR expressed in those cells (23). We have therefore repeated these experiments using PC12 cells, which express the same RyR isoform as cardiac myocytes (51, 52), and observed essentially the same result (Fig. 3). Adaptation or inactivation has been frequently invoked as an explanation for quantal Ca²⁺ release from both RyR and InsP₃R. However, it cannot explain all the quantal responses observed thus far. Indeed, for InsP₃R, the bulk of studies that have investigated a role for inactivation in generating quantal responses have concluded that it is not apparent (45-48).

The activation of both InsP₃R and RyR has been suggested to be sensitive to lumenal Ca²⁺, although this is not universally accepted, particularly for InsP₃R. The way in which lumenal Ca²⁺ may sensitize Ca²⁺ release is not fully established, but may involve direct binding to the channels (53) or allosteric effects of intermediary Ca²⁺-binding proteins such as calsequestrin (54). In the present study, increasing lumenal Ca²⁺ not only enhanced the proportion of stored Ca²⁺ that was released by a low caffeine concentration (Fig. 6C), but also decreased the caffeine concentration necessary to evoke Ca²⁺ release (Fig. 6B). The latter observation indicates that lumenal Ca²⁺ did not simply potentiate Ca²⁺ release by increasing the flux of Ca²⁺ through the RyR, and thereby enhancing Ca²⁺-induced Ca²⁺ release on the cytosolic side of the channels. Instead, lumenal Ca²⁺ altered the threshold for activation of RyR. Changing lumenal Ca²⁺ content from ~13 to 130% of the steady-state Ca²⁺ load (i.e. 10-s reloading period versus KCl treatment; Fig. 5), there was a 10-fold decrease in the threshold (from 10 to 1 mM caffeine). In a previous study, we found that augmenting the Ca²⁺ load of the intracellular stores significantly increased the frequency of spontaneous "elementary" Ca²⁺ release events in PC12 cells (55), consistent with the idea that lumenal Ca²⁺ enhances the opening of RyR. Our data also suggest that differences in the steady-state lumenal Ca²⁺ content may contribute to the normally variable sensitivity of PC12 cells to caffeine (Fig. 7).

The ability of lumenal Ca²⁺ to alter the sensitivity of RyRs to caffeine may underlie the quantal Ca²⁺ responses observed in this study. With each application of caffeine in Ca²⁺_o-free medium, there would be a decrease in lumenal Ca²⁺ content (Fig. 4), which would thereby reduce the sensitivity of RyR to caffeine so that a higher caffeine concentration would be necessary to evoke additional Ca²⁺ release. Consistent with this scheme, we observed that depletion of lumenal Ca²⁺ stores by application of caffeine in one region of a cell could globally reduce the sensitivity of RyR to that caffeine concentration (Fig. 7). Such global desensitization of caffeine responsiveness by a local caffeine application also argues against any contribution of adaptation or inactivation of RyR by cytosolic stimuli to the quantal response, since channels that were not stimulated with caffeine showed the same lack of activity as those RyR that had been stimulated.

In previous investigations of quantal Ca²⁺ release from RyR in adrenal chromaffin cells, we suggested that the concentration-dependent emptying of the caffeine-sensitive Ca²⁺ pool was due to the all-or-none emptying of functionally discrete pools (14, 17). This conclusion was largely based on the observation that treatment of cells with low caffeine concentrations in the presence of ryanodine locked a fraction of the intracellular RyR in an open subconductance state, but these constitutively open channels could not fully deplete the caffeine-sensitive Ca²⁺ pool (similar results were obtained for PC12 cells; data not shown). In undertaking the present study, we did not expect to find such a strong correlation between luminal Ca²⁺ content and caffeine sensitivity. However, in light of the results obtained, we have to revise our former conclusions and suggest that the all-or-none model does not account for quantal responses. The most plausible explanation for our previous data (14, 17) is that even RyR locked in the open state are sensitive to regulation by luminal Ca²⁺, so that they cannot release all of the intracellular caffeine-sensitive Ca²⁺ pool.

In the case of InsP₃R, permeabilized cell experiments have suggested that control of the channels by lumenal Ca²⁺ may occur over only ~30% of the possible Ca²⁺ loading range, in which case it could not explain quantal Ca²⁺ release (see Ref. 56, see also Ref. 47). However, the Ca²⁺ stores in permeabilized cells can be loaded substantially beyond the normal steady-state loading found in intact cells, in which case the "window" over which lumenal Ca²⁺ can affect channel sensitivity could appear to be insignificant. We feel it is therefore significant to show that the change in RyR sensitivity was seen over the full range of Ca²⁺ loads that can be obtained in intact PC12 cells.

Previous studies using intact cells, where lumenal Ca²⁺ was directly monitored using targeted aequorin have also pointed to a control of InsP₃R and RyR function by Ca²⁺ within the stores of various cell types (57, 58). One of the most substantial examples of changes in lumenal Ca²⁺ having a physiological consequence for RyR function was recorded using x-ray microprobe analysis in hippocampal CA3 neurons (59). Trains of action potentials enhanced the lumenal Ca²⁺ load from resting levels of ~3.5 to >72 mmol/kg dry weight. Such increases in lumenal Ca²⁺ are known to greatly enhance the activation of RyR by depolarizing stimuli, leading to substantial amplification of Ca²⁺_cyt signals by Ca²⁺-induced Ca²⁺ release (60).

In summary, our data indicate that lumenal Ca²⁺ plays a role in adjusting the sensitivity of RyR to caffeine. Indeed, differences in the steady-state Ca²⁺ load appears to underlie the variable sensitivity of the cells to caffeine. RyR adaptation does not appear to cause the quantal pattern of Ca²⁺ release. Instead, the decrease in sensitivity of Ca²⁺ release caused by declining lumenal Ca²⁺ content may be responsible for the termination of RyR activity, thereby causing graded mobilization of the Ca²⁺ pool.

	ACKNOWLEDGEMENTS

S. K. thanks Drs. K. Inoue and Y. Ohno for continuous encouragement.

	FOOTNOTES

^* This work was supported in part by the Health Science Foundation in Japan and the Biotechnology and Biological Sciences Research Council.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed: Laboratory of Molecular Signalling, The Babraham Institute, Babraham, Cambridge, CB2 4AT United Kingdom. Tel.: 44-1223-496515; Fax: 44-1223-496033; E-mail: peter.lipp@bbsrc.ac.uk.

^** Royal Society University Research Fellow.

	ABBREVIATIONS

The abbreviations used are: Ca²⁺_cyt, cytosolic Ca²⁺ concentration; ER, endoplasmic reticulum; fura-2 AM, fura-2 acetoxymethylester; InsP₃, inositol 1,4,5-trisphosphate; InsP₃R, inositol 1,4,5-trisphosphate receptor(s); RyR, ryanodine receptor(s); SR, sarcoplasmic reticulum.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES