Comparison of the Effects Exerted by Luminal Ca²⁺ on the Sensitivity of the Cardiac Ryanodine Receptor to Caffeine and Cytosolic Ca²⁺

Abstract

Ca²⁺ released from the sarcoplasmic reticulum (SR) via ryanodine receptor type 2 (RYR2) is the key determinant of cardiac contractility. Although activity of RYR2 channels is primary controlled by Ca²⁺ entry through the plasma membrane, there is growing evidence that Ca²⁺ in the lumen of the SR can also be effectively involved in the regulation of RYR2 channel function. In the present study, we investigated the effect of luminal Ca²⁺ on the response of RYR2 channels reconstituted into a planar lipid membrane to caffeine and Ca²⁺ added to the cytosolic side of the channel. We performed two sets of experiments when the channel was exposed to either luminal Ba²⁺ or Ca²⁺. The given ion served also as a charge carrier. Luminal Ca²⁺ effectively shifted the EC₅₀ for caffeine sensitivity to a lower concentration but did not modify the response of RYR2 channels to cytosolic Ca²⁺. Importantly, luminal Ca²⁺ exerted an effect on channel gating kinetics. Both the open and closed dwell times were considerably prolonged over the whole range (response to caffeine) or the partial range (response to cytosolic Ca²⁺) of open probability. Our results provide strong evidence that an alteration of the gating kinetics is the result of the interaction of luminal Ca²⁺ with the luminally located Ca²⁺ regulatory sites on the RYR2 channel complex.

Introduction

In cardiac muscle, the Ca²⁺ required for contractile activation is rapidly released from the sarcoplasmic reticulum (SR) in response to entry of small amounts of Ca²⁺ from the extracellular space to the cytosol. The key role in this process, termed Ca ²⁺ -induced Ca ²⁺ release (CICR), is played by the cardiac ryanodine receptor (RYR2)/Ca²⁺ release channel clustered in the SR membrane (Fabiato, 1985). Eighty percent of the molecular mass of the RYR2 channel is exposed to the cytosol and only 20% is embedded in membrane, indicating that some regions are accessible from the lumen. The big cytoplasmic domain is the main regulatory domain of the channel, where the binding sites for ligands such as Ca²⁺, caffeine, adenosine triphosphate (ATP), and Mg²⁺ are localized (Meissner, 1994). However, there is growing evidence that luminal regions of RYR2 channels may also contribute to the regulation of channel function. Fabiato & Fabiato (1979) demonstrated that the magnitude of CICR in skinned cardiac cells was enhanced as the Ca²⁺ loading of the SR was elevated. Fabiato (1992) later proposed that there are regulatory binding sites for Ca²⁺ in the lumen of the SR. This hypothesis focused more attention on luminal Ca²⁺ as a regulator, and experiments on different levels were performed to clarify the role of luminal Ca²⁺ in the regulation of CICR. Evidence from single-channel studies of RYR2 channels also points to a regulatory role for luminal Ca²⁺. It has been consistently shown under different experimental conditions that elevating luminal Ca²⁺ leads to an increase in RYR2 channel activity (Sitsapesan & Williams, 1994a, 1997; Lukyanenko, Györke & Györke, 1996; Györke & Györke, 1998; Xu & Meissner, 1998; Ching, Williams & Sitsapesan, 2000). The results can be interpreted on the basis of two proposed mechanisms of luminal Ca²⁺ regulation. One model suggests that Ca²⁺ flowing via the RYR2 channel activates the channel by having access to cytosolic Ca²⁺ regulatory sites (“feed-through” model). The alternative suggestion is that luminal Ca²⁺ acts at distinct sites on the luminal side of the Ca²⁺ release complex (“true luminal regulation” model). The feed-through model is favored by Xu & Meissner (1998). In support of this concept, they revealed a close correlation between the effects of luminal Ca²⁺ on RYR2 activity and the magnitude of Ca²⁺ flux from lumen to cytosol. However, the aforementioned findings do not rule out the possibility that luminal Ca²⁺ acts on the luminal side of the RYR2 channel. This issue was addressed directly by Györke & Györke (1998), who performed measurements at high membrane potentials, when the electrochemical gradient did not support lumen-to-cytosol Ca²⁺ fluxes, and at high cytosolic Ca²⁺ concentrations in order to saturate cytosolic Ca²⁺ activating sites. In their experiments, luminal Ca²⁺ still potentiated native canine RYR2 channels, thus supporting the existence of luminally located Ca²⁺ regulatory sites. The subsequent work of Ching et al. (2000) provided further strong evidence in favor of the true luminal regulation model. The ability of native RYR2 channels to respond to luminal Ca²⁺ was altered after the channels were exposed to luminal trypsin. Apparently, some of the luminal Ca²⁺ regulatory sites were damaged. The source of the aforementioned discrepancies is unclear, but they might reflect differences in the degree of protein purification. It has been pointed out that harsher purification procedures make the channel more prone to the feed-through action of luminal Ca²⁺ (Györke et al., 2004). Purification can result in a loss of some regulatory proteins involved in luminal Ca²⁺ sensing or damage of regulatory Ca²⁺ binding sites located directly on the luminal side of the RYR2 channel. Consistent with this hypothesis, a complex of calsequestrin, triadin 1 and junctin was found to communicate changes in luminal Ca²⁺ to the RYR2 channel (Györke et al., 2004). Triadin 1 and junctin anchor calsequestrin, which serves as the actual luminal Ca²⁺ sensor to the RYR2 channel. At low luminal Ca²⁺ concentrations, calsequestrin inhibited activity of the RYR2 channel, whereas elevating the luminal Ca²⁺ led to a dissociation of calsequestrin from the triadin 1- junction complex and thus a gradual activation of the channel. However, this finding did not exclude the possibility that luminal Ca²⁺ exerts activation effects by binding directly to the luminal aspect of the RYR2 channel. Indeed, a detailed study on the skeletal isoform of RYR channels (RYR1) showed that an increase in the luminal Ca²⁺ concentration caused a biphasic increase in channel activity. An initial, rapid phase was fully reversible and therefore attributed to a direct effect of luminal Ca²⁺. A second, slow phase was interpreted as the result of calsequestrin dissociation from the channel because full recovery required the addition of excess calsequestrin (Beard et al., 2002; Beard, Laver & Dulhunty, 2004). A similar study should be conducted for the RYR2 channel to test whether the reported effect of luminal Ca²⁺ is partially or fully mediated by the action of calsequestrin.

Most reports on the effect of luminal Ca²⁺ on RYR2 channels have focused on determination of the differences in channel activity induced by elevating luminal Ca²⁺, whereas a detailed examination of channel gating kinetics has not been of particular interest. Györke & Györke (1998) performed a partial analysis showing that luminal Ca²⁺ enhanced channel activity, primarily by increasing the number of openings. In contrast, Xu & Meissner (1998) found that the effect of luminal Ca²⁺ was manifested predominantly by increased open dwell times, while the number of events remained unchanged. This inconsistency is likely to be attributable to the application of different types of cytosolic activators (caffeine, ATP) that exert their own effects on gating kinetics. To contribute to a deeper understanding of the mechanism by which luminal Ca²⁺ regulates the RYR2 channel, we investigated its effects on the caffeine and cytosolic Ca²⁺ sensitivity of a single RYR2 channel, using the planar lipid bilayer method. Two sets of experiments were performed under asymmetric conditions using either 53 mM luminal Ba²⁺ or 53 mM luminal Ca²⁺ as a charge carrier. Furthermore, we examined the effects of luminal Ca²⁺ on the gating kinetics of the RYR2 channel over the whole range of channel activity. Our results suggest that luminal Ca²⁺ exerts its effect on the RYR2 channel by enhancing its sensitivity to cytosolic activators such as caffeine. Moreover, the open and closed dwell times were considerably prolonged over the whole range (response to caffeine) or the partial range (response to cytosolic Ca²⁺) of open probability. Our results indicate that the effect on gating kinetics is likely attributed to the action of luminal Ca²⁺ on the luminal face of the channel, providing further evidence in support of the existence of distinct intraluminal Ca²⁺ sensing sites which regulate the behavior of the RYR2 channel.

Materials and Methods

PREPARATION OF SR MEMBRANE VESICLES

Cardiac SR microsomes were isolated from Wistar rat heart according to Buck, Lachnit & Pessah (1999) with a few modifications. The isolated left ventricles from four hearts (4 g) were homogenized with a Tissue Tearor (Biospec Products, Inc., Bartlesville, OK, USA) in 5 volumes of homogenization buffer (1 M KCl, 10 mM Tris-maleate) and a cocktail of protease inhibitors (Roche Applied Science, Mannheim, Germany; 1 mM benzamidine, 5 μg/ml pepstatin, 5 μg/ml leupeptin, 1 μM calpain inhibitor I, 1 μg/ml aprotinin, 1 mM 4-(2-Aminoethyl)-benzenesolfonyl. The homogenate was centrifuged for 20 min at 10,000 × g _max at 4°C. The supernatant was discarded, and the remaining pellet was homogenized in an ice-cold homogenization buffer and centrifuged for 20 min at 6,000 × g _max at 4°C. The supernatant was centrifuged for 25 min at 24,000 × g _max at 4°C, and the resulting supernatant was further centrifuged for 120 min at 41,000 × g _max at 4°C. The final pellet was resuspended in resuspension buffer: 10% sucrose, 10 mM Tris-maleate, 0.9% NaCl (pH 6.8). Aliquots were snap-frozen in liquid N₂ and stored at −70°C until used.

SINGLE-CHANNEL RECORDINGS

RYR2 channels were incorporated into a planar lipid bilayer and single-channel currents recorded under voltage-clamp conditions. Cardiac SR microsomes were added to the cis chamber near the planar lipid bilayer formed from a 3:1 mixture of DOPE/DOPS (Avanti Polar Lipids, Alabaster, AL) across a 150 μm aperture in the wall of a polystyrene cup. Fusion of microsomes was promoted by KCl added to the cis chamber (corresponding to cytosol). After incorporation of a Ca²⁺ channel, the KCl gradient was eliminated by perfusion of the cis chamber with cis solution (10 ml). The cis chamber was filled with 1 ml of 250 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 125 mM Tris, 50 mM KCl, 1 mM ethyleneglycoltetraacetic acid (EGTA) and 0.5 or 0.36 mM CaCl₂ (pH 7.35); and the trans chamber (corresponding to lumen) was filled with 1 ml of 53 mM Ca(OH)₂, 50 mM KCl, 250 mM HEPES or 53 mM Ba(OH)₂, 50 mM KCl and 250 mM HEPES (pH 7.35). The free Ca²⁺ concentration was calculated by WinMaxc32 version 2.50 (http://www.stanford.edu/∼cpatton/maxc.html). In cytosolic Ca²⁺/Ba²⁺ competition experiments, various mixtures of 1 mM EGTA, 1 mM ethylenediaminetetraacetic acid (EDTA) and 1 mM N-Carboxylmethyl-N-(2-hydroxyethyl)-N,N′-ethylenediglycine (HEDTA) were used to buffer both cytosolic Ca²⁺ and Ba²⁺. To investigate the possibility that luminal Ba²⁺ modulates RYR2 channel activity by direct binding to the luminal face of the channel, experiments were conducted in asymmetric Cs⁺ solutions (cis solution = 400 mM CsCH₃O₃S, 10 mM HEPES, 1 mM EGTA, 0.5 mM CaCl₂, 8 mM caffeine [pH 7.35]; trans solution = 20 mM CsCH₃O₃S, 10 mM HEPES [pH 7.35]). Cs⁺ was used as a charge carrier. The trans chamber was connected to the head-stage input of a Warner BC-525D amplifier (Warner Instruments, Hamden, CT), and the cis chamber was held at ground. The holding potential was 0 mV in all experiments. Electrical signals were filtered at 1 kHz, digitized at 4 kHz and analyzed. Data acquisition and analysis were performed with a commercially available software package (pCLAMP 5.5; Axon Instruments, Burlingame, CA) using an IBM-compatible Pentium computer and an A/D-D/A converter (Digidata 1322A, Axon Instruments). The open probability (P _o) was calculated from continuous records of >2-min duration. For the purpose of gating kinetics analysis, the records were divided into 30-s intervals for luminal Ba²⁺ and 60-s intervals for luminal Ca²⁺. The analyzed intervals were prolonged for luminal Ca²⁺ due to an insufficient number of events collected on shorter intervals. The average open and closed times were calculated on these intervals as a standard arithmetic average. The resulting values for luminal Ca²⁺ and Ba²⁺ were further averaged on the defined intervals of P _o and statistically compared. At P _o ∼ 0.5, the open and closed time histograms were constructed and fit by a sum of exponential curves, yielding values of the mean open and closed times. The results are reported as average ± standard deviation (SD). The significance of differences was analyzed by the Student t-test with Welch’s approximation and statistical significance was accepted at P < 0.05.

Results

SR microsomes were fused into a planar lipid membrane, and single-channel currents through RYR2 channels were recorded under asymmetric conditions with either luminal Ca²⁺ or luminal Ba²⁺ as a charge carrier. The net current at 0 mV membrane potential was in the lumen-to-cytosol (trans to cis) direction, and the magnitude was constant. Experiments with luminal Ba²⁺ were considered to serve as a control. To exclude the potential masking effects of contaminating Ca²⁺, 1 mM EGTA was added to the trans solution containing Ba(OH)₂. To establish that experiments with luminal Ba²⁺ can indeed be used as a control, we tested whether gradual elevation of luminal Ba²⁺ concentration up to 53 mM results in changes in channel activity. Experiments were performed in the absence of luminal Ca²⁺, and current recordings were obtained in asymmetric 400/20 mM (cis/trans) Cs⁺ solutions at 0 mV holding potential. When luminal Ba²⁺ concentration was <40 mM, the primary charge carrier was Cs⁺ and a cytosol-to-lumen ion flux through the channel pore was favored. However, elevation of luminal Ba²⁺ concentration from 40 to 53 mM led to a change in the current direction. Ba²⁺ became the primary charge carrier defining the current direction from lumen to cytosol. To clearly see a potential decrease or increase in channel activity, RYR2 channels were activated by ∼8 mM caffeine to P _o ∼ 0.5. Free cytosolic Ca²⁺ concentration was kept constant (100 nM). Under these experimental conditions, we did not observe any significant activation or inhibition of the RYR2 channel when Ba²⁺ concentration on the luminal side was varied over a range from 0 to 53 mM (data not shown).

EFFECTS OF LUMINAL CA²⁺ ON CAFFEINE-ACTIVATED RYR2 CHANNELS

Caffeine has been widely used as a pharmacological tool for releasing Ca²⁺ from intracellular stores in the study of excitation-contraction coupling (Fabiato & Fabiato, 1976; Endo & Kitazawa, 1978; Song et al., 2002; Yoshihara et al., 2005). In our work, we used caffeine as a probe to clarify the way in which luminal Ca²⁺ affects the function of RYR2 channels. All of the experiments were conducted at ∼90 nM cytosolic Ca²⁺ when RYR2 channels exhibited low activity regardless of whether Ca²⁺ or Ba²⁺ was present on the luminal side of the channel. In Figure 1A, a single RYR2 channel was recorded in the presence of various concentrations of caffeine using either luminal Ba²⁺ (upper traces) or luminal Ca²⁺ (lower traces) as a charge carrier. An elevation of the caffeine concentration resulted in the sequential activation of the channel approximately to its maximum extent under both conditions. Figure 1B summarizes these experiments by plotting P _o against caffeine concentration for each experiment. Individual curves were fit with the Hill equation, and the values of EC₅₀ (concentration of caffeine needed to achieve half-maximal activation) were further averaged. It is apparent that in the presence of luminal Ca²⁺ the EC₅₀ was significantly reduced from 7.94 ± 0.69 mM (n = 5) to 1.76 ± 1.25 mM (n = 8) (P < 0.01) (Fig. 1C). The maximal extent of channel activation by caffeine was not affected by luminal Ca²⁺. P _{o max} was 0.74 ± 0.17 (n = 5) for luminal Ba²⁺ and 0.87 ± 0.10 (n = 8) for luminal Ca²⁺ (Fig. 1C). Our results support and complement the finding of Xu & Meissner (1998), who kept the concentration of caffeine constant and increased the luminal Ca²⁺ concentration.

Figure 1.

Effects of luminal Ca²⁺ on the response of the RYR2 channel to caffeine. (A) Representative current traces of single RYR2 channel at varying concentrations of caffeine applied from the cytosolic side of the channel in the presence of either luminal Ba²⁺ (upper traces) or luminal Ca²⁺ (lower traces). Channel openings are in the upward direction. Dashes at the left of the tracings indicate closed state of the channel (C). (B) The relationship between normalized P _o and caffeine concentration is displayed for each analyzed experiment. Single-channel activities were determined at 0 mV potential and under asymmetric conditions using either luminal Ba²⁺ (●) or luminal Ca²⁺ (○) as a charge carrier. (C) Statistical comparison of average EC₅₀ and P _{o max} for caffeine response of RYR2 channel recorded at either luminal Ba²⁺ or Ca²⁺. The graphs show that luminal Ca²⁺ shifted significantly the EC₅₀ to lower values (1.76 ± 1.25 mM for luminal Ca²⁺ vs. 7.94 ± 0.69 mM for luminal Ba²⁺, **P < 0.01) but did not exert any effect on the maximal extent of channel activation (0.87 ± 0.10 for luminal Ca²⁺ vs. 0.74 ± 0.17 for luminal Ba²⁺). Data are presented as average ± SD. More than four experiments were used to calculate average values.

An additional source that can contribute to elucidating the molecular mechanism of luminal Ca²⁺ action is the analysis of channel gating kinetics. In the presence of low caffeine and cytosolic Ca²⁺ concentrations, an insufficient number of open and closed events for constructing and fitting dwell time histograms was collected. Therefore, as a first step, we determined the average dwell times calculated as a standard arithmetic average (Fig. 2).

Figure 2.

Gating kinetics of caffeine-activated RYR2 channel is modified by luminal Ca²⁺. Average open time, closed time and frequency of opening accumulated from 30-s recordings for luminal Ba²⁺ (●) and 60-s recordings for luminal Ca²⁺ (○) were further averaged on defined intervals of P _o and compared. Statistically significant differences between luminal Ba²⁺ and Ca²⁺ (P < 0.05) are shown for all tested intervals of P _o except the first interval for the average closed time vs. P _o dependence and the first and the last intervals for the frequency of opening vs. P _o dependence. Error bars represent SD.

It is apparent from the representative current traces (Figure. 1A) that luminal Ca²⁺ influenced not only the EC₅₀ for caffeine activation but also the duration of open and closed events. Figure 2 summarizes the results of gating kinetics analysis. The average open and closed times and the frequency of opening as an additional parameter describing channel behavior were plotted against P _o. In both cases, for luminal Ba²⁺ and luminal Ca²⁺, the activation of the channel by caffeine was manifested by a prolongation of the average open time, a shortening of the average closed time and a bell-shaped dependence of the frequency of opening on P _o. However, in the presence of luminal Ca²⁺, the RYR2 channel exhibited a much lower frequency of opening over the whole range of P _o. The maximum value reached for luminal Ba²⁺ was 65 ± 16 Hz (n = 4) and that for luminal Ca²⁺ was 22 ± 7 Hz (n = 5). Luminal Ca²⁺ effectively slowed down the channel gating kinetics, inevitably resulting in significant prolongation of average open and closed times in order to reach a similar degree of activation as for luminal Ba²⁺. This conclusion was confirmed by statistical comparison of gating kinetics parameters over the whole range of channel activity (Fig. 2).

At P _o ∼ 0.5, the frequency of opening reached a maximum; therefore, a sufficient number of events for the construction and fitting of dwell time distributions was collected. For luminal Ba²⁺, the open and closed time distributions were well fit by the sum of three exponential curves, demonstrating the occupancy of at least three different open and three different closed states. The mean dwell time and corresponding occupancy (n = 4) for each open/closed component are listed in Table 1. Similarly, luminal Ca²⁺ produced a pattern of gating which was also characterized with at least three open and three closed states (Table 1). Each of the three open and two of three closed times (τ_C2 and τ_C3) determined for luminal Ca²⁺ were significantly longer than for luminal Ba²⁺. Furthermore, the channel exposed to luminal Ca²⁺ preferentially occupied the states with the two longest open times (τ_O2 and τ_O3) in contrast to luminal Ba²⁺, where the states with the two shortest open times were preferred (τ_O1 and τ_O2). A similar tendency as in the case of the open states was found for the occupancy of closed states in the presence of luminal Ba²⁺, and some minor differences were revealed for luminal Ca²⁺ when the channel preferentially occupied the state with the longest and the shortest closed times (τ_C1 and τ_C3). Taken together, luminal Ca²⁺ significantly prolonged the mean open and closed times and, in general, shifted the occupancy in favor of states with longer dwell times.

Table 1. Effects of luminal Ca²⁺ on mean dwell times of caffeine-activated RYR2 channel determined at Po ∼ 0.5

The enhanced sensitivity of the RYR2 channel to caffeine might be explained by both of the suggested mechanisms of luminal Ca²⁺ regulation. To separate them, we performed an additional set of experiments in the absence of luminal Ca²⁺ to eliminate its direct effects on the luminal face of the channel. Furthermore, an accumulation of Ca²⁺ ions near the cytosolic domain of the channel was mimicked by elevating the cytosolic Ca²⁺ concentration above basal level (∼90 nM). Under these conditions, we tested the sensitivity of the RYR2 channel to caffeine. EC₅₀ values were plotted against the corresponding cytosolic Ca²⁺ concentrations (Fig. 3A). Cytosolic Ca²⁺ (173 nM) was able to shift the EC₅₀ for caffeine sensitivity to a similar concentration as was found for luminal Ca²⁺ (EC₅₀ = 1.57 ± 1.00 mM, n = 4; EC₅₀ = 1.76 ± 1.25 mM, n = 8, respectively). Furthermore, we compared the gating kinetics obtained from this set of experiments with those determined for the caffeine-activated channel exposed to luminal Ca²⁺ and 90 nM cytosolic Ca²⁺. The average open and closed times were significantly shorter, and the frequency of opening was significantly higher over the whole range of P _o (Fig. 3B). Again, at P _o ∼ 0.5, we performed a detailed analysis of gating kinetics by fitting open and closed time histograms by the sum of three exponential curves. The resulting parameters are listed in Table 2. All of the three open times and two of the three closed times were significantly shorter. Furthermore, the occupancy of open and closed states characterized by the two shortest times (τ_O1, τ_O2 and τ_C1, τ_C2) was preferred. Whereas the shift in EC₅₀ of caffeine sensitivity might be interpreted in terms of both proposed mechanisms of luminal Ca²⁺ regulation, the outcomes of a gating kinetics analysis are difficult to reconcile with the feed-through model.

Figure 3.

Effects of cytosolic Ca²⁺ on caffeine response of RYR2 channel exposed to luminal Ba²⁺. (A) Relationship between EC₅₀ for caffeine activation and cytosolic Ca²⁺ concentration for RYR2 channel. Data points are displayed as average ± SD from three or four experiments. (B) Average open time, closed time and frequency of opening accumulated from 30-s recordings for luminal Ba²⁺ (●) and 60-s recordings for luminal Ca²⁺ (○) were further averaged on defined intervals of P _o and compared. Statistically significant differences between luminal Ba²⁺ and Ca²⁺ (P < 0.05) are shown for all tested intervals of P _o except the first interval for the average closed time vs. P _o dependence and the first and last intervals for the frequency of opening vs. P _o dependence. Error bars represent SD.

Table 2. Mean dwell times of the RYR2 channel exposed to luminal Ba²⁺ and coactivated by cytosolic Ca²⁺ and caffeine determined at Po ∼ 0.5

LUMINAL CA²⁺ AND ITS IMPACT ON CYTOSOLIC CA²⁺ -ACTIVATED RYR2 CHANNELS

In most studies, the activation effect of luminal Ca²⁺ was not seen when solely Ca²⁺ was used as a cytosolic activator of RYR2 channels (Sitsapesan & Williams, 1994a, 1997; Györke & Györke, 1998). We reexamined this conclusion under our experimental conditions and performed a detailed gating kinetics analysis. This set of experiments was conducted in the absence of caffeine, and RYR2 channels were activated solely by cytosolic Ca²⁺. Recordings at five different cytosolic steady-state Ca²⁺ concentrations are shown for both luminal Ba²⁺ (Fig. 4A, upper traces) and luminal Ca²⁺ (Fig. 4A, lower traces). All channels exhibited very low activity at <0.1 μM Ca²⁺ and reached maximal activation at 0.5 μM Ca²⁺. The relationships between P _o and the cytosolic Ca²⁺ concentration (Fig. 4B) were analyzed using the Hill equation. These analyses yielded similar EC₅₀ values of 0.18 ± 0.04 μM (n = 10) for luminal Ba²⁺ and 0.19 ± 0.04 μM (n = 11) for luminal Ca²⁺. Likewise, the extent of maximal activation was not altered (0.94 ± 0.08 [n = 10] for luminal Ba²⁺ and 0.97 ± 0.07 [n = 11] for luminal Ca²⁺, Fig. 4C). A detailed analysis of the gating kinetics profile of Ca²⁺-activated RYR2 channels was one of the main aims of our study. When no flux of Ca²⁺ via channel in the lumen-to-cytosol direction is allowed (luminal Ba²⁺), an increase in P _o to ∼0.5 was due to the remarkable increase in the frequency of opening (Fig. 5). A reduction in the average closed time was found mainly to account for this observed increase in P _o during activation. A further decrease in the frequency of opening for P _o > 0.5 was mediated by a considerable increase in the average open time. An interesting behavior was revealed for the channel exposed to luminal Ca²⁺. For P _o < 0.5, the frequency of opening was very low in comparison with luminal Ba²⁺, thus resembling the gating kinetics of caffeine-activated RYR2 channels. Surprisingly, at P _o ∼ 0.5, the channel switched to a different mode of gating and started to behave as in the presence of luminal Ba²⁺. This change is documented by a considerable increase in the frequency of opening and a decrease in the average open and closed times. At P _o > 0.5, no significant differences in the examined parameters for luminal Ca²⁺ and Ba²⁺ were detected. This observation was further supported with the fitting of dwell time distributions by the sum of three exponential curves. At P _o ∼ 0.5, no changes were found between luminal Ca²⁺ and Ba²⁺ in either magnitude of dwell time or occupancy of the corresponding open and closed states (Table 3).

Figure 4.

Luminal Ca²⁺ does not alter response of RYR2 channel to cytosolic Ca²⁺. (A) Representative current traces for RYR2 channel activated by cytosolic Ca²⁺ recorded under experimental conditions when either luminal Ca²⁺ (upper traces) or luminal Ba²⁺ (lower traces) was present. Channel openings are in the upward direction. Dashes at the left of the tracings indicate closed state of the channel (C). The relationship between P _o and cytosolic Ca²⁺ concentration is shown in (B). Data points displayed are individual P _o measurements from more than nine experiments. Single-channel activities were determined at 0 mV membrane potential and under asymmetric conditions using either luminal Ba²⁺ (●) or luminal Ca²⁺ (○) as a charge carrier. (C) Statistical analysis revealed that neither the EC₅₀ (0.19 ± 0.04 μM for luminal Ca²⁺ vs. 0.18 ± 0.04 μM for luminal Ba²⁺) nor the P _{o max} was modified by luminal Ca²⁺ (0.97 ± 0.07 for luminal Ca²⁺ vs .0.94 ± 0.08 for luminal Ba²⁺). Data are presented as average ± SD.

Figure 5.

Luminal Ca²⁺ exerts effect on gating kinetics of cytosolic Ca²⁺-activated RYR2 channel with P _o < 0.5. Average values for open time, closed time and frequency of opening accumulated from 30-s recordings for luminal Ba²⁺ (●) and 60-s recordings for luminal Ca²⁺ (○) were further averaged on defined intervals of P _o and compared. Statistically significant differences between luminal Ba²⁺ and Ca²⁺ (P < 0.05) are shown for P _o < 0.5 except the first P _o interval for the frequency of opening vs. P _o dependence. For P _o > 0.5, differences were not statistically significant. Error bars represent SD.

Table 3. Mean dwell times of cytosolic Ca²⁺-activated RYR2 channel determined at Po ∼ 0.5 are not altered by luminal Ca²⁺

COMPETITION BETWEEN CA²⁺ AND BA ²⁺ FOR CYTOSOLIC REGULATORY SITES OF RYR2 CHANNELS

Knowledge about the interaction of Ba²⁺ and RYR2 channels is limited only to the findings that Ba²⁺ is not able to activate the channel from the cytosolic side (Liu, Pasek & Meissner, 1998). There is missing information about the ability of Ba²⁺ to compete with Ca²⁺ for cytosolic activation sites. However, this information is crucial for the correct interpretation of our results. To address this issue, we examined the sensitivity of RYR2 channels to cytosolic Ca²⁺ at various concentrations of cytosolic Ba²⁺. This set of experiments was performed in the presence of luminal Ba²⁺ to avoid the direct effects of luminal Ca²⁺ on the channel properties. The determined values of EC₅₀ were plotted against the concentration of cytosolic Ba²⁺ (Fig. 6). Surprisingly, we found that Ba²⁺ is an effective competitor, with the onset of competition at ∼30 μM (Fig. 6). Due to the fact that there was not a difference in EC₅₀ for the sensitivity of RYR2 channels to cytosolic Ca²⁺ examined for luminal Ba²⁺ and luminal Ca²⁺, we assumed that the Ba²⁺ passing the channel could temporarily increase the Ba²⁺ concentration near the cytosolic face of the channel up to 30 μM. Considering this finding, we examined whether 30 μM cytosolic Ba²⁺ is able to shift the EC₅₀ of caffeine activation. The experiments were done in the presence of luminal Ca²⁺ to eliminate the flux of Ba²⁺ to cytosol and, thus, to keep the concentration of cytosolic Ba²⁺ constant. The RYR2 channel exposed to cytosolic Ba²⁺ was activated by caffeine with an EC₅₀ of 1.85 ± 1.50 mM (n = 6), which is comparable to the EC₅₀ determined in the absence of cytosolic Ba²⁺ (1.76 ± 1.25 mM, n = 8). Thus, this indicates that it is not essential to consider the competition effect between Ca²⁺ and Ba²⁺ to cytosolic regulatory sites in the interpretation of observed differences in the values of EC₅₀ for caffeine activation of the RYR2 channel exposed to either luminal Ca²⁺ or luminal Ba²⁺.

Figure 6.

Competition effect between Ba²⁺ and Ca²⁺ for cytosolic regulatory sites on the RYR2 channel. Relationship between the EC₅₀ for cytosolic Ca²⁺ activation of RYR2 channel and cytosolic Ba²⁺ concentration. Various mixtures of EDTA, EGTA and HEDTA were used to chelate both cytosolic Ca²⁺ and Ba²⁺. The onset of competition appeared at ∼30 μM of cytosolic Ba²⁺. Data points are displayed as average ± SD from five to seven experiments.

Discussion

In the present study, the effects of luminal Ca²⁺ on caffeine and cytosolic Ca²⁺ activation of RYR2 channels were investigated. We already know that the RYR2 channel is activated by increasing the concentration of luminal Ca²⁺. What is clear from the present study is that luminal Ca²⁺ exerts noticeable effects also on channel gating kinetics. We were inspired by the work of Sitsapesan & Williams (1994a), Györke & Györke (1998) and Xu & Meissner (1998). All three groups performed the single-channel measurements and systematically manipulated the concentration of luminal Ca²⁺ while the concentration of cytosolic activators was maintained constant (sulmazol, ATP, caffeine). We decided to conduct complementary experiments, keeping luminal Ca²⁺ concentration constant and changing the concentration of caffeine and Ca²⁺ applied to the cytosolic face of the RYR2 channel. We conducted experiments at either 53 mM Ca²⁺ or 53 mM Ba²⁺ on the luminal face of the channel to strengthen the proposed effects of luminal Ca²⁺. At the same time, chosen ions served as charge carriers.

Experiments with luminal Ba²⁺ mimicked the situation when no Ca²⁺ is present on the luminal face of the channel. They were considered to be a control on the following evidence. Firstly, we found that luminal Ba²⁺ as high as 53 mM was not able to significantly activate or inhibit caffeine-activated RYR2 channels when asymmetric Cs⁺ solutions were used. Our experiments were performed in the absence of luminal Ca²⁺ to avoid any competition between luminal Ca²⁺ and Ba²⁺ for both luminal and cytosolic regulatory sites on the RYR2 channel. However, the competition between Ca²⁺ and Ba²⁺ was studied by Tripathy & Meissner (1996). They reported that the addition of 1–5 mM Ba²⁺ to the luminal side of the RYR1 channel (skeletal muscle isoform) caused a significant decrease in P _o due to competition between Ba²⁺ and Ca²⁺ for binding sites in the channel pore and on the cytosolic face. When Ca²⁺ flux from the lumen was eliminated by appropriate holding potential, no activation or inhibition was observed when luminal Ba²⁺ was added. Importantly, this finding supports our result. It seems that under these condition, Ba²⁺ did not compete with Ca²⁺ for luminally located regulatory sites. The possible explanation could be that 50 μM luminal Ca²⁺ is much lower than the K _D for Ca²⁺ binding to the luminal face of the RYR1 channel. Thus, Ba²⁺ had access to unoccupied binding sites, and this situation was similar to our finding. Secondly, Ba²⁺ is not able to activate RYR2 channels from the cytosolic side (Liu et al., 1998). Thirdly, we suggest that, although Ba²⁺ competes with Ca²⁺ for cytosolic regulatory sites, the observed difference in the EC₅₀ for caffeine activation of RYR2 channels exposed to either luminal Ba²⁺ or luminal Ca²⁺ is presumably not attributed to this competition effect. Luminal Ba²⁺ did not alter sensitivity of the RYR2 channel to cytosolic Ca²⁺, indicating that accumulation of Ba²⁺ near the cytosolic face due to flux of Ba²⁺ in the lumen-to-cytosol direction did not exceed 30 μM concentration when the onset of competition between Ca²⁺ and Ba²⁺ for cytosolic regulatory sites appeared. Importantly, the EC₅₀ for caffeine activation was not affected by 30 μM cytosolic Ba²⁺. Considering the fact that the sensitivity of the RYR2 channel to caffeine is regulated by subactivating concentrations of cytosolic Ca²⁺ (Sitsapesan & Williams, 1990), we can conclude that luminal Ba²⁺ permeating the channel does not cause the shift in the EC₅₀ for caffeine activation to a higher concentration due to the competition with Ca²⁺ for cytosolic regulatory sites but, rather, that luminal Ca²⁺ sensitized the RYR2 channel to caffeine in respect to luminal Ba²⁺. Fourthly, the value of EC₅₀ for caffeine activation of RYR2 channels exposed to luminal Ba²⁺ obtained in our study is similar to values observed in the absence of luminal Ba²⁺ and Ca²⁺ (Xu & Meissner, 1998; Lokuta et al., 1997). At present, it is impossible to do precise quantitative comparisons of our and already published parameters of the gating kinetics for RYR2 channels due to missing detailed information about the gating kinetics profile in the absence of luminal Ba²⁺ and Ca²⁺. Despite this, a simple coarse comparison showed us that there is some trend of a slight prolongation of the mean and average open and closed times by luminal Ba²⁺ (Sitsapesan & Williams, 1994b; Györke & Györke, 1998). Thus, it is unlikely that luminal Ba²⁺ exerts a significant effect on the channel gating kinetics and the remarkable prolongation of average and mean dwell times by luminal Ca²⁺ is obviously related to the presence of Ca²⁺ on the luminal face of the channel.

Our study led to new observations that may have important implications for understanding the principles of the mechanisms underlying the regulation of RYR2 channels by luminal Ca²⁺.

First, the sensitivity of RYR2 channels to caffeine was enhanced when luminal Ca²⁺ was present. Moreover, the gating kinetics of caffeine-activated RYR2 channels exposed to luminal Ca²⁺ was remarkably slowed down. Although the shift in EC₅₀ to a lower concentration can also be induced in the absence of luminal Ca²⁺ by direct addition of Ca²⁺ to cis solution (cytosolic face of the channel), remarkable prolongation in dwell times typical for luminal Ca²⁺ was not induced by this maneuver. We found that even a small increase in cytosolic Ca²⁺ from 90 to 173 nM was enough to shift the EC₅₀ of caffeine activation. Importantly, 173 nM cytosolic Ca²⁺ alone did not induce considerable activation of the channel but, rather, sensitized the channel to caffeine (Sitsapesan & Williams, 1990).

We simulated the potential accumulation of luminal Ca²⁺ near the cytosolic face of the channel due to Ca²⁺ flow from the lumen by simple addition of cytosolic Ca²⁺ to the cis solution. We are aware that our steady-state experimental model only provides a course approximation to what actually occurs in the close vicinity of the cytosolic face. According to Xu & Meissner (1998), luminal Ca²⁺ has access to cytosolic regulatory sites only during channel opening. When the channel closes, Ca²⁺ accumulated near the cytosolic face rapidly dissipates and the channel feels only the equilibrated concentration of cytosolic Ca²⁺ in the bulk. This hypothesis was further supported by findings of Laver, O’Neill & Lamb (2004). They used the competition between Mg²⁺ and Ca²⁺ for cytosolic regulatory sites of the RYR1 channel (skeletal muscle isoform) as an elegant probe for the ionic environment near the cytosolic mouth of the channel. The lack of expected competition between cytosolic Mg²⁺ and luminal Ca²⁺ for a single channel was suggested to result from limited temporal access between cytosolic and luminal solutions. Interestingly, the presence of competition for raft of RYR1 channels demonstrated that the cytosolic regulatory sites of the closed channel can still have access to luminal Ca²⁺ via the adjacent open channel.

In our experimental model, cytosolic Ca²⁺ concentration mimicking the accumulation of luminal Ca²⁺ on the cytosolic face is constant regardless of the channel state. However, according to mathematical models, it could be a relevant situation (Mazzag, Tignanelli & Smith, 2005). It was predicted that Ca²⁺ released by the open channel (residual Ca²⁺) may influence subsequent channel gating. This effect depends on the time scale for Ca²⁺ domain formation and collapse near the cytosolic face compared to the characteristic time scale for channel gating. When the localized Ca²⁺ domain forms and collapses slowly, the cytosolic regulatory binding sites for Ca²⁺ experience essentially the same concentration of cytosolic Ca²⁺ regardless of channel state (Mazzag et al., 2005). Our experimental conditions (∼4 pA Ca²⁺ current in the lumen-to-cytosol direction, slow cytosolic Ca²⁺ chelating by 1 mM EGTA) favor slow generation and dissipation of Ca²⁺ gradients, supporting the validity of our experimental model.

In summary, we can conclude that at least alteration of the gating kinetics of the caffeine-activated RYR2 channel is accounted for by binding of Ca²⁺ to distinct luminally located regulatory sites.

Secondly, consistent with published findings (Sitsapesan & Williams, 1994a; Györke & Györke, 1998), the values of EC₅₀ for cytosolic Ca²⁺ activation were not significantly different when the RYR2 channel was exposed to either luminal Ba²⁺ or Ca²⁺. We can state that if there was a global rise in cytosolic Ca²⁺ to 173 nM due to accumulation of ions passing the channel, we would be able to see the shift in EC₅₀ to a lower concentration under our experimental conditions. For this purpose, we extended the tested interval of cytosolic Ca²⁺ concentration down to 50 nM to get a definable baseline for dose response. The lack of any effect on the EC₅₀ has been previously interpreted in terms of both the true luminal regulation (Sitsapesan & Williams, 1997) and feed-through (Laver et al., 2004) models. Sitsapesan & Williams (1997), preferring the true luminal regulation model, hypothesized that the conformation of the channel activated solely by cytosolic Ca²⁺ is such that the regulatory sites for Ca²⁺ on the luminal face of the RYR2 channel to which luminal Ca²⁺ binds in the presence of a second cytosolic ligand are inaccessible. However, we revealed that luminal Ca²⁺ modified the gating kinetics of cytosolic Ca²⁺-activated RYR2 channel, thus providing evidence that luminally located binding sites for Ca²⁺ are also accessible when cytosolic Ca²⁺ is used as a sole activator. Experiments with luminal Ba²⁺ defined the activation effect of Ca²⁺ acting solely on the cytosolic side of the channel; therefore, they were crucial for appropriate interpretation of our results. When the channel exposed to luminal Ca²⁺ exhibited low activity (P _o < 0.5), the gating kinetics was very slow and resembled the behavior of the caffeine-activated channel exposed to luminal Ca²⁺. When the activity was enhanced (P _o > 0.5) by increasing the cytosolic Ca²⁺ concentration, the channel started to open and close more frequently, resembling the behavior of the RYR2 channel exposed to luminal Ba²⁺. If the lumen-to-cytosol Ca²⁺ flux was required for this effect, we would expect to see a faster gating kinetics profile also for P _o < 0.5, inasmuch as such gating kinetics is typical for the channel exposed to luminal Ba²⁺ and activated solely by cytosolic Ca²⁺. Furthermore, the remarkable prolongation of average dwell times for P _o < 0.5 is more likely accounted for by binding of Ca²⁺ to distinct luminally located regulatory sites. Considering this conclusion, we can suggest that cytosolic Ca²⁺ is a specific activator of the RYR2 channel due to its ability to take control of the channel gating, at least for P _o > 0.5. In contrast, caffeine does not exhibit a similar potency as cytosolic Ca²⁺ to compete with luminal Ca²⁺ for regulation of the gating behavior.

In contrast, Laver et al. (2004) favored the feed-through model to interpret the findings that the activation of RYR2 channels by luminal Ca²⁺ required the presence of cytosolic activators other than Ca²⁺. Considering the fact that binding of Ca²⁺ to cytosolic regulatory sites must take place for transition of the resting RYR2 channel to an open state, it has been hypothesized that luminal Ca²⁺ emanating from the channel pore does not augment channel opening because the cytosolic regulatory sites are already occupied. According to our analysis, the action of luminal Ca²⁺ should not be hampered by cytosolic Ca²⁺. It is reasonable to assume that the channel does not recognize whether Ca²⁺ was added directly to the cytosolic side or flowed from the luminal side. When the channel opens, Ca²⁺ flux in the lumen-to-cytosol direction will cause a temporal increase in the local concentration of Ca²⁺ near the already occupied cytosolic regulatory sites. Under these conditions, dissociation/association of Ca²⁺ from or to the channel will be more frequent, leading to gradual channel activation as observed for the RYR2 channel activated solely by cytosolic Ca²⁺.

Obviously, in order to clearly distinguish the effects attributed to the action of luminal Ca²⁺ on either the luminal or cytosolic face of the RYR2 channel, it is necessary to find an experimental approach when only one mechanism will operate. Testing the mutated RYR2 channels resistant to cytosolic Ca²⁺ activation could be the next step (Li & Chen, 2001).

In summary, we demonstrate for the first time that luminal Ca²⁺ modifies the gating kinetics of RYR2 channels manifested by a significant prolongation of the open and closed event durations over the whole range of P _o. Importantly, this effect is presumably related to the action of luminal Ca²⁺ on the luminal side of the channel. Whether the alteration of gating kinetics induced by luminal Ca²⁺ is the result of calsequestrin dissociation from the triadin 1–junctin complex or if it is attributed to the binding of Ca²⁺ to regulatory sites localized directly on the luminal part of the RYR2 channel remains to be clarified. In general, our results favor the true luminal regulation model and provide further evidence of the existence of functional Ca²⁺ regulatory sites on the luminal side of the RYR2 channel.