INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IN CARDIAC EXCITATION-CONTRACTION (E-C) coupling, release of Ca²⁺ from the sarcoplasmic reticulum (SR) occurs by Ca²⁺-induced Ca²⁺ release (CICR; Refs. 6-8), triggered largely by Ca²⁺ entering the cell through sarcolemmal L-type voltage-dependent Ca²⁺ channels (4, 5, 15). At the macroscopic level, the SR Ca²⁺ release flux is found to be smoothly graded and roughly proportional to the L-type Ca²⁺ current (I_Ca) despite the positive feedback implicit in the CICR mechanism. The leading hypothesis to explain this "paradox of control" is the "local control theory" (23), which proposes that the trigger for CICR from a given SR Ca²⁺ release channel (commonly known as ryanodine receptor; RyR) is the stochastic local Ca²⁺ microdomain created by nearby L-type channels at the diad junction. According to this theory, the graded control of macroscopic SR Ca²⁺ release is actually achieved by statistical recruitment of individual, stochastic release events, each of which may be locally regenerative.

One implication of this theory is that the macroscopic SR Ca²⁺ release flux may be affected quite differently by those interventions that alter the unitary L-type Ca²⁺ current (i_CaL) or open duration of the L-type channel than by those that alter its frequency of opening, even if the effect on macroscopic I_Ca is the same. One manifestation of this difference, which has been observed experimentally, is that the "gain" of CICR, i.e., ratio of SR Ca²⁺ release flux to I_Ca, decreases with increasing membrane voltage (26) because as the Ca²⁺ reversal potential is approached, the decline in i_CaL reduces the efficiency of activation of RyRs, which probably depends cooperatively on (local) Ca²⁺ concentration ([Ca²⁺]). The theory also predicts that SR Ca²⁺ release flux should eventually saturate as I_Ca is increased by raising extracellular [Ca²⁺] (Ca_o²⁺), affecting primarily i_CaL, or by prolonging the duration of L-type channel openings. The present studies were implemented to seek evidence of such saturation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell isolation. Single myocytes were enzymatically dispersed from rat heart ventricles by Langendorff perfusion with collagenase and protease in the presence of low Ca²⁺ by use of procedures described in detail previously (12, 22).

Solutions. Myocytes were bathed in a modified HEPES solution that contained (in mM) 137.0 NaCl, 15 glucose, 1.3 MgSO₄, 1.0 CaCl₂, 20 HEPES, and 0.02 TTX (to block Na⁺ channels). In addition, KCl was replaced with CsCl (10 mM), and 4-aminopyridine (5 mM) was added to block K⁺ currents. pH was adjusted to 7.4 with NaOH. Two different test solutions were rapidly applied from a micropipette to the cells: 1) CaCl₂ concentration was increased from 1.0 to 4.0 mM, or 2) 1 µM FPL-64176, an L-type Ca²⁺ channel agonist, was added to the control medium. The patch-pipette solution contained (in mM) 20 HEPES, 1 MgCl₂, 5 MgATP, and 0.05 indo 1 (pentapotassium salt). For inhibition of K⁺ currents, K⁺ was replaced with CsCl (120 mM) and tetraethylammonium (20 mM). To suppress Ca²⁺ influx via the Na⁺/Ca²⁺ exchanger, Na⁺ was omitted. pH was adjusted to 7.2 with CsOH. In some experiments with FPL-64176, 20 mM EGTA was added to the patch-pipette solution to abolish electrically stimulated cytosolic [Ca²⁺] (Ca_i²⁺; intracellular [Ca²⁺]) transients and to minimize the effect of Ca_i²⁺ on L-type Ca²⁺ channels.

Electrophysiological measurements. The experiments were carried out at room temperature (22-24°C) on a modified Zeiss inverted microscope (IM-35) equipped for simultaneous recordings of indo 1 fluorescence and membrane current as described elsewhere (22). Membrane currents were measured under whole cell voltage-clamp control (10) with a patch-clamp circuit (Axopatch 1-D). Patch-type pipettes had a tip resistance of 1.5-3.5 M when filled with the cell dialyzing solution. In most instances, the amplitude of I_Ca was determined as the difference between peak inward current and the current measured at the end of the 100-ms voltage step. In experiments using FPL-64176, which markedly slowed I_Ca inactivation, I_Ca amplitude was assessed as the difference between peak inward current during exposure to the compound and current measured at the end of the voltage step before exposure to FPL-64176 and/or after its washout. The latter closely corresponded to steady current measured at the end of 100-ms voltage pulses measured after Cd²⁺-dependent complete abolition of I_Ca, in the presence or absence of FPL-64176, in cells dialyzed with EGTA (see Fig. 3).

Measurements of Ca_i²⁺. Ca_i²⁺ activity, measured with indo 1 (pentapotassium salt), was calculated from an in vivo calibration curve after subtraction of the background fluorescence at each fluorescence emission wavelength, after correction by an algorithm that takes into account the intensity fluctuations of the xenon strobe lamp used for fluorescence excitation (22). The background fluorescence was measured after establishment of a "giga-seal" before breaking into the cell. To reduce the effects of photon noise, the data were filtered by convolution with a Gaussian function; one-half width at the 1/e point was 18 ms.

Experimental protocol. The myocytes were stimulated at 0.5 Hz, with 100-ms voltage-clamp steps to 0 mV, from a holding potential of 75 mV (in the presence of 20 µM TTX). Rapid exposures to a higher (4 mM) Ca_o²⁺ or to 1 µM FPL-64176 were abruptly commenced during an interstimulus interval by rapid exchange of the control solution around individual cells with a test solution applied from a micropipette in a manner similar to a concentration-clamp system described previously (16) that, reportedly, replaced the control solution with a test solution in <100 ms. Although interpretation of the present results is not dependent on the rapidity and/or completeness of replacement of the control medium with the test solutions, it is of note that abrupt exposures to a higher Ca_o²⁺ applied during an interstimulus interval augmented peak I_Ca to virtually the same extent as longer (60 s) steady-state exposures to 4 mM Ca_o²⁺ (Fig. 1).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effects of selective augmentation of L-type Ca2+ current (ICa) and/or sarcoplasmic reticular (SR) Ca2+ loading on Ca2+ release. Original recordings showing simultaneously measured cytosolic Ca2+ (Cai2+) transients (top), their maximal rate of rise (dCai2+/dtmax, where t is time; inset), and ICa elicited by 100-ms voltage-clamp pulses to 0 mV, from a holding potential of 75 mV, applied at 0.5 Hz (with Na+ and K+ currents and Ca2+ influx via Na+/Ca2+ exchanger blocked) in a representative rat ventricular myocyte dialyzed with 50 µM indo 1 (pentapotassium salt). A: ICa and Cai2+ transient during a steady-state control depolarization (C) are superimposed on those during the first depolarization after an abrupt increase in extracellular Ca2+ concentration (Cao2+; from 1 to 4 mM) applied during an interstimulus interval. The higher Cao2+ markedly augments peak ICa but has a disproportionately smaller effect on amplitude of corresponding Cai2+ transient or its dCai2+/dtmax. VM, membrane voltage. B: steady-state beat measured during 60-s exposure to 4 mM Cao2+ shows an increase in peak ICa similar to that in A and a large augmentation of the Cai2+ transient amplitude and its dCai2+/dtmax (inset), completed during a markedly prolonged time to peak vs. control and/or vs. the initial depolarization in higher Cao2+. C: peak ICa returns to control levels on the first depolarization after switch back to control Cao2+, whereas amplitude and dCai2+/dtmax (inset) of Cai2+ transients remain markedly potentiated vs. control.

Data analysis. In each cell, control I_Ca and Ca_i²⁺ activity were averaged in five steady-state beats before exposures to a higher Ca_o²⁺ or FPL-64176. In electrically stimulated cells, extended (60 s) exposures to a higher Ca_o²⁺ or FPL-64176 produced spontaneous oscillations of Ca_i²⁺ (attributable to SR Ca²⁺ overload), which reduced the amplitude of I_Ca and/or Ca_i²⁺ transients during the subsequent beat. Thus "steady-state" effects of exposures to a higher Ca_o²⁺ or FPL-64176 on I_Ca and Ca_i²⁺ activity were assessed by averaging five selected beats, which were not preceded by spontaneous Ca_i²⁺ oscillations during the interstimulus interval. In each myocyte, steady-state control values were compared with I_Ca and Ca_i²⁺ activity measured during the first depolarization after an abrupt exposure to a higher Ca_o²⁺ or FPL-64176, applied during an interstimulus interval, or with I_Ca and Ca_i²⁺ activity measured in response to the first depolarization after an abrupt switch from 4 to 1 mM Ca_o²⁺, when I_Ca amplitude decreased to (or below) the control level. The mean percentage of change was calculated as the average of percentage change in each cell. Student's t-test was used to determine the statistical significance of experimental interventions. Data are reported as means ± SE for n number of cells. A value of P < 0.05 was considered statistically different.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of selective, voltage-independent augmentation of peak I_Ca on SR Ca²+ release. We first examined a pure effect of an augmentation of peak I_Ca (i.e., in the presence of a constant voltage-dependent activation of L-type Ca²⁺ channels and the SR Ca²⁺ loading) on the SR Ca²⁺ release, indexed as change in the maximal rate of rise of Ca_i²⁺ (dCa_i²⁺/dt_max, a sensitive index of the SR Ca²⁺ release flux; Refs. 20, 21, 26) or the change in the amplitude of the Ca_i²⁺ transient (systolic-diastolic; Ca_i²⁺). Voltage-independent augmentation of I_Ca was effected by two methods: 1) an abrupt increase in Ca_o²⁺ (Fig. 1A) or 2) an abrupt exposure to FPL-64176 (see Fig. 4A), which rapidly increased peak I_Ca or its slowly inactivating component, respectively, in the absence of changes in the SR Ca²⁺ loading.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Pooled data from experiments of the type shown in Fig. 1 showing relative changes (%control) in ICa amplitude, dCai2+/dtmax and amplitude of simultaneously measured Cai2+ transient, and the gain of ICa-dependent SR Ca2+ release, assessed as dCai2+/dtmax/ICa or as change in () Cai2+/ICa, measured during the first depolarization after an abrupt increase in Cao2+ (A), steady depolarizations in higher Cao2+ (B), and the first depolarization after washout of high Cao2+ (C), which reduced peak ICa to or below control levels measured before switch from 1 to 4 mM Cao2+. Note difference in scale of the vertical axis in B vs. A and C. Data are mean values ± SE from 6 cells. Significantly different from control: *P < 0.05, **P < 0.01, and ***P < 0.001. NS, not significant.

Effect of combined actions of augmented I_Ca and the resulting augmentation of the SR Ca²+ loading on Ca²⁺ release. Figure 1B shows that prolonged exposure to the higher Ca_o²⁺ (for an additional 30 depolarizations/60 s) did not further increase I_Ca or alter steady-state diastolic Ca_i²⁺ but greatly augmented the amplitude and dCa_i²⁺/dt_max (Fig. 1B, inset) of the steady-state Ca_i²⁺ transient (cf. assessment of steady-state responses in MATERIALS AND METHODS). Interestingly, the times to peak dCa_i²⁺/dt_max and to peak Ca_i²⁺ were markedly prolonged. Compared with the control conditions, the average augmentation of the dCa_i²⁺/dt_max and the amplitude of the Ca_i²⁺ transients were 71 ± 10% (P < 0.001) and 81 ± 11% (P < 0.001), respectively (Fig. 2B). The CICR gain was increased by 19 ± 5% (P < 0.02) when assessed as dCa_i²⁺/dt_max/I_Ca and by 27 ± 6% (P < 0.01) when assessed as Ca_i²⁺/I_Ca (Fig. 2B). The time to peak dCa_i²⁺/dt_max was prolonged by 37 ± 5% (from 20 ± 1 to 28 ± 2 ms, P < 0.01), and the time to peak Ca_i²⁺ was prolonged by 85 ± 14% (from 37 ± 3 to 69 ± 6 ms, P < 0.002). An intriguing, novel observation provided by the experiments in Figs. 1B and 2B is that, in addition to the well-established positive inotropic effect of a higher Ca_o²⁺ on the Ca_i²⁺ transient amplitude (2, 25), combined actions of a larger I_Ca trigger and an augmented SR Ca²⁺ loading, achieved by the prolonged exposure to higher Ca_o²⁺, markedly prolong the SR Ca²⁺ release flux, as indicated by a significant increase in the time to peak dCa_i²⁺/dt_max and a prolonged time to peak Ca_i²⁺ (see also Fig. 4B).

Effect of selective augmentation of SR Ca²+ loading on I_Ca-dependent Ca²⁺ release. To assess a pure effect of an augmentation of the SR Ca²⁺ loading on CICR, we switched back to the control (1 mM) Ca_o²⁺ and measured the amplitude and dCa_i²⁺/dt_max of the initial Ca_i²⁺ transient elicited by an I_Ca of the same (or smaller) amplitude than in the control, before exposure to a higher Ca_o²⁺. Return of the control amplitude of I_Ca typically occurred within one to two depolarizations (2-4 s) after a switch to the control Ca_o²⁺, so that a decrease in the SR Ca²⁺ load (vs. steady state in 4 mM Ca_o²⁺) was presumably small. Figure 1C shows a marked residual potentiation of the amplitude of the Ca_i²⁺ transient and its dCa_i²⁺/dt_max versus control, accompanied by a shift in the time to peak Ca_i²⁺ or dCa_i²⁺/ dt_max toward control values. On average, during the first depolarization after washout of a higher Ca_o²⁺, the amplitude of I_Ca was virtually the same as in control (97 ± 2% of control, P > 0.05), whereas dCa_i²⁺/dt_max and the amplitude of the corresponding Ca_i²⁺ transients remained markedly increased, by 37 ± 8% (P < 0.005) and 45 ± 11% (P < 0.01), respectively (Fig. 2C). Accordingly, the CICR gain assessed as dCa_i²⁺/dt_max/I_Ca was increased by 41 ± 8% (P < 0.005), and Ca_i²⁺/I_Ca was increased by 49 ± 11% (P < 0.01) versus control, i.e., they were 20% larger than the CICR gain during maximally potentiated, steady-state depolarizations measured in the continuous presence of a higher Ca_o²⁺ (Fig. 2, C vs. B). However, the time to peak dCa_i²⁺/dt_max and to peak Ca_i²⁺ averaged 21 ± 2 and 50 ± 6 ms, respectively, i.e., not statistically different from control.

Effect of FPL-64176 on I_Ca in adult rat ventricular myocytes dialyzed with EGTA. FPL-64176, a novel L-type Ca²⁺ channel agonist, is thought to increase both the probability of channel opening and the mean channel opening time. Effects of FPL-64176 on I_Ca have been initially characterized in neonatal rat ventricular myocytes (18) and GH₃ cells (14) dialyzed with 15 mM EGTA and superfused with solutions containing 10 mM Ba²⁺ or 10 mM Ca²⁺ as charge carriers. Thus it was of interest to first examine the effect of FPL-64176 on I_Ca in adult rat ventricular myocytes dialyzed with EGTA (20 mM) and superfused with a more physiological (1 mM) Ca_o²⁺ (Fig. 3). In another set of experiments (Fig. 4), we examined the effect of an augmented Ca²⁺ influx during the slowly inactivating component of I_Ca on the SR Ca²⁺ release in cells not dialyzed with EGTA.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effects of FPL-64176 (FPL, indicated by *) on voltage dependence of Ca2+ influx via Cd2+-sensitive L-type Ca2+ channels in rat ventricular myocytes dialyzed with EGTA. A: original recordings showing superimposed tracings of Cd2+-sensitive L-type Ca2+ current, measured under control conditions and after a 60-s exposure to 1 µM FPL-64176, in a representative rat ventricular myocyte dialyzed with 20 mM EGTA during steps to test potentials of 30 to 0 mV from a holding potential of 75 mV (with Na+ and K+ currents and Ca2+ influx via Na+/Ca2+ exchanger blocked). FPL-64176 markedly slows inactivation of ICa during voltage pulses and of the "tail" ICa, elicited on repolarization. Notably, increasing step potential progressively reduced the effect of FPL-64176 on peak ICa, which, at steps to greater than or equal to 10 mV, was similar in magnitude to that elicited in control. B: superimposed tracings of membrane current measured in same cell at steps from 75 to 0 mV in absence and presence of FPL-64176 and after exposure to 100 µM Cd2+, a specific blocker of L-type Ca2+ channels. C: Cd2+-sensitive L-type Ca2+ current in control (CTRL; left) and in presence of FPL-64176 (right) as assessed from same tracings as in B by electronic subtraction of Cd2+-insensitive component of membrane current.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Effects of FPL-64176 on ICa and SR Ca2+ release. A: rapid exposure to 1 µM FPL-64176, applied during an interstimulus interval, markedly augments Ca2+ influx during the slowly inactivating component of ICa but has virtually no effect on the dCai2+/dtmax (inset) or amplitude of the corresponding Cai2+ transient. C, control. B: a longer exposure to FPL-64176 (for 60 s) produces a further attenuation of ICa inactivation and augmentation of Ca2+ influx during the slowly inactivating component of ICa and the tail ICa and a large increase in dCai2+/dtmax and the Cai2+ transient amplitude (note 2-fold difference in Cai2+ scales between A and B), achieved in presence of markedly prolonged times to peak dCai2+/dtmax (inset) and to peak Cai2+ vs. control.

Effect of selective augmentation of the slowly inactivating component of I_Ca on Ca²+ release. Figure 4A shows a representative example of the effect of an abrupt exposure to 1 µM FPL-64176 during an interstimulus interval on I_Ca and the corresponding Ca_i²⁺ transient. In this and three other experiments of that type, FPL-64176 did not significantly increase peak I_Ca but markedly slowed its inactivation, thus augmenting the amount of Ca²⁺ entering the cell via I_Ca (I_Ca integral) during the voltage pulse and via the tail I_Ca activated on repolarization. However, dCa_i²⁺/dt_max (Fig. 4A, inset) and the amplitude of the Ca_i²⁺ transients, and/or their time to peak, were virtually unaltered during the first depolarization after an abrupt exposure to FPL-64176 (Fig. 4A) despite a large increase in the I_Ca integral (on average by 117 ± 8%, mean ± SE; n = 4, P < 0.001). Because in the present experiments we were primarily concerned with the initial Ca²⁺ release, the effect of FPL-64176 on the tail I_Ca (which increased by severalfold) or on secondary SR Ca²⁺ release(s) was not systematically examined.

Effect of combined actions of augmented Ca²+ influx during slowly inactivating component of I_Ca and the resulting augmentation of the SR Ca²⁺ loading on Ca²⁺ release. Figure 4B shows that a longer exposure to FPL-64176 (for an additional 30 depolarizations/60 s) further delays I_Ca inactivation, thus increasing its integral versus control, on average, by 229 ± 10% (n = 4, P < 0.001) and the integral of the tail I_Ca 10- to 15-fold. Although steady-state diastolic Ca_i²⁺ was not altered, the amplitude and dCa_i²⁺/dt_max (Fig. 4B, inset) of the steady-state Ca_i²⁺ transients were greatly augmented, on average, by 204 ± 33% (P < 0.01) and 93 ± 19% (P < 0.02), respectively. The time to peak maximal dCa²⁺/dt was prolonged by 63 ± 5% (from 20 ± 2 to 32 ± 3 ms, P < 0.01), and time to peak Ca_i²⁺ was prolonged by 124 ± 33% (from 35 ± 4 to 75 ± 5 ms, P < 0.002). The most important observation in these experiments is that, in addition to the expected inotropic effect of FPL-64176 on the dCa_i²⁺/dt_max and amplitude of the Ca_i²⁺ transients, a combination of maximally potentiated SR Ca²⁺ loading and increased Ca²⁺ influx via I_Ca also prolongs the SR Ca²⁺ release flux, as indicated by a significant increase in the time to peak dCa_i²⁺/dt_max and prolonged time to peak Ca_i²⁺ (Fig. 4B).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

It is now generally accepted that nanoscale Ca²⁺ domains play an important role in the communication between the L-type Ca²⁺ channel and the RyR(s) that leads to release of SR Ca²⁺ in the heart (1). This implies that different interventions that increase the ensemble average I_Ca are not equivalent, because the number of Ca²⁺ channels, their unitary current, and their open duration each play a distinct role in the local CICR process. In this paper, we have examined the effects of two interventions on I_Ca characteristics: increasing extracellular Ca²⁺, which is expected to increase i_CaL, and the Ca²⁺ channel agonist FPL-64176, which acts mainly by prolonging channel openings. It is necessary, when altering I_Ca, to distinguish between two different kinds of effects: those due to alteration of the trigger for CICR and those due to alteration of SR Ca²⁺ loading. After acute, persistent change in I_Ca, SR Ca²⁺ content changes relatively slowly, over a number of beats (9). More importantly, in the present experiments, exposures to a higher Ca_o²⁺ or FPL-64176 were rapidly effected during the interval between steady-state voltage-clamp pulses to 0 mV, from a holding potential of 75 mV (with fast Na⁺ current and Ca²⁺ influx via the sarcolemmal Na⁺/Ca²⁺ exchanger blocked). This experimental design most likely prevented any appreciable changes in the SR Ca²⁺ content before first depolarization after a rapid switch to the test solution (Figs. 1A and 4A). Thus we were able to separate out the effect of a selective, pure increase in the I_Ca trigger on SR Ca²⁺ release in the presence of virtually unaltered SR Ca²⁺ content. Because CICR is highly dependent on changes in the SR Ca²⁺ loading (2, 5, 11, 12), an incomplete separation of the effect on I_Ca from that on SR Ca²⁺ content in experiments of the type shown in Figs. 1A and 4A would abate the observed disparity between changes in I_Ca and the magnitude of the corresponding Ca²⁺ SR release. The present experiments show that increasing the i_CaL trigger by abruptly raising Ca_o²⁺ produces only a small increment in the SR Ca²⁺ release flux during a depolarization to 0 mV (Figs. 1A and 2A), whereas increasing Ca²⁺ content of the SR by repeated stimulation at high Ca_o²⁺ can produce a large effect (Figs. 1B and 2B). This implies that under the conditions of these experiments, the SR Ca²⁺ release at 0 mV was saturated with respect to the i_CaL trigger. In other words, the unitary trigger current at this voltage is not the rate-limiting step; rather, some other link in the E-C coupling/CICR pathway limits Ca²⁺ release. Two obvious links are the status and number of RyRs and SR Ca²⁺ load. If i_CaL at 0 mV is sufficient to guarantee the maximal activation of a release unit of RyRs whenever an L-type channel opening occurs, a further increase in the unitary current will not recruit additional SR Ca²⁺ release units. However, the scenario assumes that increased Ca²⁺ influx via L-type channels cannot activate release units other than the one to which a given L-type channel is coupled. Our results are not sufficient to prove that this is the situation, because another powerful competition process is present: depletion of SR Ca²⁺ during the same beat. A single depolarization to 0 mV is sufficient to release 40-60% of SR Ca²⁺ (2). Therefore, the driving force for SR Ca²⁺ release will decline substantially during a given depolarization, even if the state of SR Ca²⁺ loading at the start of the depolarization is controlled by means of conditioning pulses, as in the present study. The recruitment of additional SR release units will hasten this decline, so that the total amount of Ca²⁺ released from the SR will increase in a less than proportional manner. These two kinds of saturation (i.e., of RyR activation or of SR Ca²⁺ availability) are, in principle, distinguishable kinetically (e.g., SR Ca²⁺ depletion during the beat should have a more marked effect on total Ca²⁺ release than on maximum rate of release, and no effect on initial rate of release). However, theoretical separation of these two kinds of saturation relies heavily on the parameters chosen within the framework of a numerical model for the local control of Ca²⁺, and theoretical separation is not reliable on the basis of the experiments herein.

A more surprising result of the present experiments was the finding that the time to peak of dCa_i²⁺/dt_max or Ca_i²⁺ is prolonged after prolonged exposure to higher Ca_o²⁺. This suggests that the duration of RyR release flux is prolonged. Under these conditions, both the I_Ca trigger and the SR Ca²⁺ content are increased. Increasing either SR Ca²⁺ content or i_CaL might be expected to shorten the duration of SR Ca²⁺ release, because both would increase the initial rate of release of a limited pool of SR Ca²⁺, assuming that unitary Ca²⁺ flux via RyRs (i_RyR) is roughly proportional to SR luminal [Ca²⁺], which is thought to increase at a greater rate than the SR Ca²⁺ content because of calsequestrin saturation (2). Additionally, the increased local [Ca²⁺] produced by increasing either i_CaL or i_RyR would be expected to shorten individual release events if the latter are terminated by putative Ca²⁺-dependent inactivation of the RyR. The paradoxical increase in the SR Ca²⁺ release duration that we observed might be produced by at least two mechanisms. First, it is possible that under near-physiological conditions, i_RyR is actually saturated as a function of the SR Ca²⁺ content. A larger SR Ca²⁺ load would then lead to a prolongation of release, at a fixed maximum release rate, if SR Ca²⁺ depletion were to contribute significantly to the termination of release (see above). A second possibility is that the prolonged release comes from corbular SR. In the rat ventricle, ~40% of the SR Ca²⁺ release terminals are corbular, i.e., located away from any junction with the sarcolemma (13). Release of Ca²⁺ from corbular SR must be triggered by the global rather than local (at the mouth of L-type Ca²⁺ channel in the diadic cleft) Ca_i²⁺ increase. Because the latter is much smaller, and increases much more slowly than the local [Ca²⁺] in the diadic cleft, corbular SR might contribute a slow release component seen only under conditions of both high SR Ca²⁺ loading and a larger I_Ca trigger (Figs. 1B, 2B, and 4B). Another possibility that must be kept in mind when interpreting the results of experiments in which SR Ca²⁺ loading is varied is that SR luminal [Ca²⁺] might directly modulate the gating properties of the RyRs. An elucidation of this possibility is not approachable experimentally.

The observed effects of FPL-64176 on SR Ca²⁺ release are consistent with those of increasing Ca_o²⁺, i.e., that SR Ca²⁺ release is saturated with respect to both the unitary amplitude and duration of individual local Ca²⁺ trigger events. The fact that peak I_Ca during a depolarization to 0 mV is not greatly increased by FPL-64176 indicates that its effects on unitary properties of the L-type Ca²⁺ channel in cardiac myocytes are complex. These effects will need to be more fully characterized at the single-channel level before the effect of FPL-64176 on the gain of CICR can be fully understood. It is noteworthy, however, that a decrease in CICR gain has been observed with the dihydropyridine Ca²⁺ channel agonist BAY K 8644 (17), which acts at a different site on the L-type Ca²⁺ channel than FPL-64176 (18).

The interpretation of the present results, that the E-C coupling pathway under the present conditions saturates with respect to the unitary properties of the L-type Ca²⁺ current trigger, leads to important implications for models of the E-C coupling process. In local control theories (23, 24), the regulation of SR Ca²⁺ release by the I_Ca trigger occurs mainly by way of recruitment of release units of RyRs. Saturation of this process in the present experiments implies that most RyR units are being activated by a depolarization to 0 mV in the presence of 1 mM Ca_o²⁺. This has important implications for the mechanism of termination of macroscopic SR Ca²⁺ release (19). Full activation of the SR Ca²⁺ release units under control conditions is also, itself, somewhat paradoxical. Current dogma holds that both L-type Ca²⁺ channels and RyRs are present in numbers greatly in excess of what would be required to produce the maximum observed rate of SR Ca²⁺ release (3). The present results, therefore, suggest that the unitary properties of these channels, particularly the RyRs, may be quite different in situ than in isolation.