The effect of heart rate on the membrane currents of isolated sheep Purkinje fibres

The control of cardiac ventricular excitability by autonomic pathways

Central to the genesis of ventricular cardiac arrhythmia are variations in determinants of excitability. These involve individual ionic channels and transporters in cardiac myocytes but also tissue factors such as variable conduction of the excitation wave, fibrosis and source-sink mismatch. It is also known that in certain diseases and particularly the channelopathies critical events occur with specific stressors. For example, in hereditary long QT syndrome due to mutations in KCNQ1 arrhythmic episodes are provoked by exercise and in particular swimming. Thus not only is the static substrate important but also how this is modified by dynamic signalling events associated with common physiological responses. In this review, we examine the regulation of ventricular excitability by signalling pathways from a cellular and tissue perspective in an effort to identify key processes, effectors and potential therapeutic approaches. We specifically focus on the autonomic nervous system and related signalling pathways.

Ca(2+) current facilitation is CaMKII-dependent and has arrhythmogenic consequences

The cardiac voltage gated Ca²⁺ current (I_Ca) is critical to the electrophysiological properties, excitation-contraction coupling, mitochondrial energetics, and transcriptional regulation in heart. Thus, it is not surprising that cardiac I_Ca is regulated by numerous pathways. This review will focus on changes in I_Ca that occur during the cardiac action potential (AP), with particular attention to Ca²⁺-dependent inactivation (CDI), Ca²⁺-dependent facilitation (CDF) and how calmodulin (CaM) and Ca²⁺-CaM dependent protein kinase (CaMKII) participate in the regulation of Ca²⁺ current during the cardiac AP. CDI depends on CaM pre-bound to the C-terminal of the L-type Ca²⁺ channel, such that Ca²⁺ influx and Ca²⁺ released from the sarcoplasmic reticulum bind to that CaM and cause CDI. In cardiac myocytes CDI normally pre-dominates over voltage-dependent inactivation. The decrease in I_Ca via CDI provides direct negative feedback on the overall Ca²⁺ influx during a single beat, when myocyte Ca²⁺ loading is high. CDF builds up over several beats, depends on CaMKII-dependent Ca²⁺ channel phosphorylation, and results in a staircase of increasing I_Ca peak, with progressively slower inactivation. CDF and CDI co-exist and in combination may fine-tune the I_Ca waveform during the cardiac AP. CDF may partially compensate for the tendency for Ca²⁺ channel availability to decrease at higher heart rates because of accumulating inactivation. CDF may also allow some reactivation of I_Ca during long duration cardiac APs, and contribute to early afterdepolarizations, a form of triggered arrhythmias.

Accumulation of slowly activating delayed rectifier potassium current (IKs) in canine ventricular myocytes

In guinea-pig ventricular myocytes, in which the deactivation of slowly activating delayed rectifier potassium current (IKs) is slow, IKs can be increased by rapid pacing as a result of incomplete deactivation and subsequent current accumulation. Whether accumulation of IKs occurs in dogs, in which the deactivation is much faster, is still unclear. In this study the conditions under which accumulation occurs in canine ventricular myocytes were studied with regard to its physiological relevance in controlling action potential duration (APD). At baseline, square pulse voltage clamp experiments revealed that the accumulation of canine IKs could occur, but only at rather short interpulse intervals (< 100 ms). With action potential (AP) clamp commands of constant duration (originally recorded at rate of 2 Hz), an accumulation was only found at interpulse intervals close to 0 ms. Transmembrane potential recordings with high-resistance microelectrodes revealed, however, that at the fastest stimulation rates with normally captured APs (5 Hz) the interpulse interval exceeded 50 ms. This suggested that no IKs accumulation occurs, which was supported by the lack of effect of an IKs blocker, HMR 1556 (500 nM), on APD. In the presence of the beta-adrenergic receptor agonist isoproterenol (isoprenaline, 100 nM) the accumulation with AP clamp commands of constant duration was much more pronounced and a significant accumulating current was found at a relevant interpulse interval of 100 ms. HMR 1556 prolonged APD, but this lengthening was reverse rate dependent. AP clamp experiments in a physiologically relevant setting (short, high rate APs delivered at a corresponding rate) revealed a limited accumulation of IKs in the presence of isoproterenol. In conclusion, a physiologically relevant accumulation of IKs was only observed in the presence of isoproterenol. Block of IKs, however, led to a reverse rate-dependent prolongation of APD indicating that IKs does not have a dominant role at short cycle lengths.

Calcium, calmodulin, and calcium-calmodulin kinase II: heartbeat to heartbeat and beyond

Calcium (Ca) is the key regulator of cardiac contraction during excitation-contraction (E-C) coupling. However, differences exist between the amount of Ca being transported into the myocytes upon electrical stimulation as compared to Ca released from the sarcoplasmic reticulum (SR). Moreover, alterations in E-C coupling occur in cardiac hypertrophy and heart failure. In addition to the direct effects of Ca on the myofilaments, Ca plays a pivotal role in activation of a number of Ca-dependent proteins or second messengers, which can modulate E-C coupling. Of these proteins, calmodulin (CaM) and Ca-CaM-dependent kinase II (CaMKII) are of special interest in the heart because of their role of modulating Ca influx, SR Ca release, and SR Ca uptake during E-C coupling. Indeed, CaM and CaMKII may be associated with some ion channels and Ca transporters and both can modulate acute cellular Ca handling. In addition to the changes in Ca, CaM and CaMKII signals from beat-to-beat, changes may occur on a longer time scale. These may occur over seconds to minutes involving phosphorylation/dephosphorylation reactions, and even a longer time frame in altering gene transcription (excitation-transcription (E-T) coupling) in hypertrophic signaling and heart failure. Here we review the classical role of Ca in E-C coupling and extend this view to the role of the Ca-dependent proteins CaM and CaMKII in modulating E-C coupling and their contribution to E-T coupling.

Differential effects of changes in local myocardial refractoriness on atrial and ventricular latency

BACKGROUND

Assessment of myocardial refractoriness and conduction properties, critical to development and propagation of reentrant arrhythmias, is an integral part of the investigation of atrial and ventricular tachycardias through the use of programmed electrical stimulation. Local conduction itself, however, may be affected by myocardial refractoriness.

METHODS AND RESULTS

We studied the effects of changes in myocardial refractoriness on local conduction in right atrial and ventricular myocardium in 19 patients. Changes in effective, functional, and relative refractoriness were accomplished with the use of four pacing protocols, including drive train pacing cycle lengths (PCLs) of 600 and 400 milliseconds (ms) and drive train durations (DTDs) of 8 and 50 stimuli. Unipolar cathodal stimulation was performed from the distal electrode, and unipolar electrograms were recorded from the proximal three poles of quadripolar catheters with 5-mm interelectrode spaces. Atrial and ventricular effective and relative refractory periods (ERPs and RRPs) were significantly shortened by both the reduction in PCL and the increase in DTD. The reduction in the PCL shortened atrial and ventricular refractory periods significantly more than did the increase in the DTD. Changes in ventricular refractory periods were significantly greater compared with atrial refractory periods. The ratio of RRP to ERP was reduced in the atrium but significantly increased in the ventricle with reduction in PCL and increase in DTD. For all premature intervals, the conduction interval from stimulus to the first recording electrode pole was significantly greater than conduction intervals measured between subsequent electrode poles. The greatest increase in conduction interval with closely coupled premature complexes occurred between the stimulus artifact and the first recording electrode pole. Reduction in ventricular but not atrial ERP was associated with significantly increased local conduction interval.

CONCLUSIONS

First, most of the conduction delay after the stimulus artifact occurs within 5 mm from the pacing site. Second, closely coupled premature complexes delay conduction primarily by prolonging latency in the first 5 mm from the pacing site. Third, fundamental differences occur between the atrium and ventricle regarding changes in local conduction as a function of changes in ERP, suggesting that factors involved in sudden changes in refractoriness (eg, heart rate acceleration) could produce divergent effects on atrial and ventricular arrhythmogenesis.

Effects of the bradycardic agent ZD 7288 on membrane voltage and pacemaker current in sheep cardiac Purkinje fibres

The bradycardic mechanism of ZD 7288 (4-(N-ethyl-N-phenylamino)-1,2-dimethyl-6-(methylamino)pyrimidinium++ + chloride) was investigated in sheep cardiac Purkinje fibres. The pacemaker i(f)-current measured with the two-microelectrode voltage-clamp technique, as well as the diastolic depolarization rate and the frequency of spontaneously active fibres were evaluated. ZD 7288 did inhibit i(f)-current. The i(f)-amplitude recorded with a 0.8s-lasting test pulse from about -50 mV to -100 mV was reduced to 50% of control at 0.85 mumol/l and to 5% of control at 10 mumol/l. The threshold potential of i(f)-activation was unaffected at a concentration of 1 mumol/l ZD 7288. The time constant of i(f)-activation at different test potentials was not changed by 1 mumol/l ZD 7288. The drug was equally effective during i(f)-activation with a 0.5 s-lasting test pulse applied at 0.05 Hz or 0.5 Hz. During long lasting (5 s) hyperpolarizing test pulses (-120 mV) the inhibition of i(f)-current was removed. In constantly stimulated Purkinje fibres (0.5 Hz) the slope of the early diastolic depolarization was decreased by ZD 7288. The half-maximal effect occurred at 0.92 mumol/l. There was strong correlation over the concentration range of 0.01 to 10 mumol/l ZD 7288 between the decrease of the slope of early diastolic depolarization and inhibition of i(f)-amplitude recorded with 0.8s-lasting test pulses to -100 mV. The correlation coefficient was r = 0.97. These results will explain the decrease in frequency of spontaneously active (about 0.6 Hz) Purkinje fibres.(ABSTRACT TRUNCATED AT 250 WORDS)

Ultra-slow voltage-dependent inactivation of the calcium current in guinea-pig and ferret ventricular myocytes

L-type Ca2+ current, iCa, has been recorded in guinea-pig ventricular myocytes at 36 degrees C using the whole cell patch clamp technique. Intracellular Ca2+ was buffered with ethylenebis(oxonitrilo)tetraacetate (EGTA). An increase in the rate of stimulation from 0.5 to 3 Hz resulted in an abrupt decrease in iCa in the first beat at the high rate, followed by a progressive decrease (tau approx. 7 s) over the next 30 s. The changes were not the result of Ca(2+)-dependent inactivation, because similar changes occurred with either Ba2+ or Na+ as the charge carrier. During 20-s voltage clamp pulses there was an ultra-slow phase of inactivation of Ba2+ or Na+ current through the Ca2+ channel (tau approx. 6 s at 0 mV). This was confirmed by applying test pulses after conditioning pulses of different duration: the Ba2+ current during the test pulse decreased progressively when the duration of the conditioning pulse was increased progressively to 20 s. Ultra-slow inactivation of Ba2+ current was voltage dependent and increased monotonically at more positive potentials. Recovery of Ba2+ current from ultra-slow inactivation occurred with a time constant of 3.7 s at -40 mV and 0.7 s at -80 mV. The gradual decrease in iCa on increasing the rate to 3 Hz may have been the result of the development of ultra-slow voltage-dependent inactivation.

Postextrasystolic potentiation. Do we really know what it means and how to use it?

Postextrasystolic potentiation (PESP), the increase in contractility that follows an extrasystole, is an interesting phenomenon that has been known for almost 100 years. The literature on this effect is reviewed. It is found that there is significant evidence that the phenomenon is independent of muscle loading and represents a distinct property of the myocardium. Examination of the literature pertaining to the cause of the effect suggests that calcium shifts within the sarcoplasmic reticulum are responsible, although there are some conflicts with this conclusion. Regarding the utility of PESP as a diagnostic test of latent viability of ischemic myocardium, the literature review reveals contradictions and conflicts with several methodological problems of the experiments. Finally, concerning the utility of continuous PESP (paired-pacing) to augment ventricular function in the failing ventricle, the studies again are inconclusive and methodologically suspect. Conditions for the proper analysis of the PESP response are reported, and suggestions for future studies are introduced.

Analytical modeling of the hysteresis phenomenon in guinea pig ventricular myocytes

In the present study, we have demonstrated hysteresis phenomena in the excitability of single, enzymatically dissociated guinea pig ventricular myocytes. Membrane potentials were recorded with patch pipettes in the whole-cell current clamp configuration. Repetitive stimulation with depolarizing current pulses of constant cycle length and duration but varying strength led to predictable excitation (1:1) and non-excitation (1:0) patterns depending on current strength. In addition, transition between patterns depended on the direction of current intensity change and stable hysteresis loops were obtained in stimulus:response pattern vs. current intensity plots in 14 cells. Increase of pulse duration and decrease of stimulation rate contributed to a reduction in hysteresis loop areas. Changes in amplitude and shape of the subthreshold responses during the transitions from one stable pattern to the other, suggested that activity led to an increase in membrane resistance, particularly in the voltage domain between resting potential, and threshold. Therefore, we modelled the dynamic behaviour of the single cells as a function of diastolic membrane resistance, using previously published analytical solutions. Numerical iteration of the analytical model equations closely reproduced the experimental hysteresis loops in both qualitative and quantitative ways. In particular, the effect of stimulation frequency on the model was similar to the experimental findings. The overall study suggests that the excitability pattern of guinea pig ventricular myocytes accounts for hysteresis and bistabilities when current intensity is allowed to fluctuate around threshold levels.

Inhibition of potassium outward currents and pacemaker current in sheep cardiac Purkinje fibres by the verapamil derivative YS 035

The electrophysiologic mode of action and potency of the verapamil derivative YS 035 (N,N-bis-(3,4-dimethoxyphenethyl)-N-methyl amine) were investigated in sheep cardiac Purkinje fibres. Action potential duration measured at a repolarization level of -60 mV (APD-60) and membrane currents recorded with the two-microelectrode voltage-clamp technique were evaluated. At 10 mumols/l YS 035 APD-60 was increased to about 115% of reference. Prolongation measured as percentage of the respective control exhibited on the average no dependence on stimulation frequency (0.17-2 Hz). At 100 mumols/l membrane became depolarized to about -50 mV and action potentials could no longer be elicited. Further study was focussed on effects on outward currents, mostly activated at a frequency of 0.05 Hz. Transient outward current (ito) was completely blocked at 100 mumols/l and half-maximal inhibition occurred at about 14 mumols/l. Inwardly rectifying potassium current (ik1) was reduced to 47% of reference at 100 mumols/l. An initially activating outward current at positive membrane potentials (iinst) was reduced to 73% at 100 mumols/l. Time-dependent (delayed) outward current (iK) was on the average not affected up to 100 mumols/l. Besides inhibition of repolarizing outward currents YS 035 completely blocked pacemaker current (if) at 100 mumols/l and half-maximal reduction was achieved at 5 mumols/l. YS 035 (1-100 mumols/l) did not clearly affect time constants of activation at selected test potentials (IK: +35 mV; if: -90 mV) or inactivation (ito: 0 mV). Voltage-dependent control mechanisms of currents (ito, if) were not influenced by YS 035 but the amount of available current was reduced.(ABSTRACT TRUNCATED AT 250 WORDS)

Hysteresis in the excitability of isolated guinea pig ventricular myocytes

Hysteresis phenomena were demonstrated in the excitability of single, enzymatically dissociated guinea pig ventricular myocytes. Membrane potentials were recorded with patch pipettes in the whole-cell current-clamp configuration. Repetitive stimulation with depolarizing current pulses of constant cycle length and duration but varying strength led to predictable excitation (1:1) and nonexcitation (1:0) patterns depending on current strength. However, transition between patterns depended on the direction of current strength change, and stable hysteresis loops were obtained in stimulus-response pattern versus current strength plots in 31 cells. Increase of pulse duration and decrease of stimulation rate contributed to a reduction in hysteresis loop areas. In addition, at the abrupt transitions from 1:0 to 1:1 patterns, a latency adaptation phenomenon was consistently observed. Bath application of tetrodotoxin (30 microM) produced no change of hysteresis, whereas hysteresis was substantially decreased in cobalt (2 mM) superfusion experiments. Analysis of the changes in amplitude and shape of the subthreshold responses during the transitions from one stable pattern to the other suggested that activity led to an increase in membrane resistance, particularly in the voltage domain between resting and threshold potentials. We therefore modeled the dynamic behavior of the single cells, using an analytical solution aimed at calculating the recovery of activation latency as a function of diastolic membrane resistance. Numerical iteration of the analytical model equations closely reproduced the experimental hysteresis loops in both qualitative and quantitative ways. The effect of stimulation frequency on the model was similar to the experimental findings. The overall study suggests that the excitability pattern of guinea pig ventricular myocytes is responsible for hysteresis and bistabilities when current intensity is allowed to fluctuate around threshold levels.

Alpha 1-adrenoceptors reduce background K+ current in rabbit ventricular myocytes

1. Ventricular myocytes were isolated by enzymatic dispersion of adult rabbit hearts, and voltage clamped using the whole-cell variation of the patch clamp technique. Experiments were carried out at either 35 degrees C or room temperature (21-23 degrees C). 2. In the presence of 10(-3) M-4-aminopyridine to block the transient outward K+ current, and 10(-6) M-propranolol to block beta-adrenoceptors, the alpha 1-adrenergic agonist methoxamine produced action potential prolongation, and a small depolarization of the diastolic membrane potential. Under voltage clamp conditions, methoxamine decreased the magnitude of the inward rectifier K+ current, IK1, in both the inward and outward directions. This effect was dose dependent (10(-5)-10(-3) M) and fully reversible upon wash-out of the agonist. 3. The neurotransmitter noradrenaline (10(-6)-2 x 10(-5) M), in the presence of propranolol (10(-6) M), also reduced IK1 in ventricular cells, and this effect was blocked by the specific alpha 1-adrenoceptor antagonist prazosin. 4. The alpha 1-adrenoceptor-mediated decrease in IK1 in ventricular myocytes was not affected by pre-incubation of the cells with 0.5 micrograms/ml pertussis toxin (8-10 h, 30-32 degrees C). This result suggests that in rabbit ventricular cells, the alpha 1-modulation of IK1 occurs via a pertussis toxin-insensitive guanine nucleotide-binding regulatory protein. 5. These observations demonstrate that IK1 in ventricular myocytes can be modulated by cardiac alpha 1-adrenoceptors. The resulting changes in action potential repolarization and diastolic membrane potential may have significant effects on cardiac performance.

A subpopulation of cells with unique electrophysiological properties in the deep subepicardium of the canine ventricle. The M cell

Recent studies have shown that canine ventricular epicardium and endocardium differ with respect to electrophysiological characteristics and pharmacological responsiveness and that these differences are in large part due to the presence of a prominent transient outward current Ito and a spike-and-dome morphology of the action potential in epicardium but not endocardium. In attempting to quantitate these differences and assess their gradation across the ventricular wall, we encountered a subpopulation of cells in the deep subepicardial layers with electrophysiological characteristics different from those of either epicardium or endocardium. These cells, which we have termed M cells, display a spike-and-dome morphology typical of epicardium but a maximal rate of rise of the action potential upstroke that is considerably greater than that of either epicardium or endocardium. Using the restitution of the amplitude of phase 1 of the action potential as a marker for the reactivation of Ito, we showed M cells to possess a prominent 4-aminopyridine-sensitive Ito with a reactivation time course characterized by two components with fast and slow time constants. The rate dependence of action potential duration of M cells was considerably more accentuated than that of epicardium or endocardium and more akin to that of Purkinje fibers (not observed histologically in this region). Phase 4 depolarization was never observed in M cells, not even after exposure to catecholamines and/or low [K+]o. In summary, our study presents evidence for the existence of a unique subpopulation of cells in the deep subepicardium of the canine left and right ventricles with electrophysiological features intermediate between those of conducting and myocardial cells. Although their function is unknown, M cells may facilitate conduction in epicardium and are likely to influence or mediate the manifestation of electrocardiographic J waves, T waves, U waves, and long QT intervals and contribute importantly to arrhythmogenesis.

Rate dependence of action potential duration and refractoriness in canine ventricular endocardium differs from that of epicardium: role of the transient outward current

Previous studies have provided evidence for an important contribution of the transient outward current to the electrical activity of canine ventricular epicardium, but not endocardium. The present study examines the characteristics of action potential duration and refractoriness in these two tissue types. The time and rate dependence of changes in action potential duration and refractoriness observed in epicardium were significantly more accentuated than in endocardium. The restitution of action potential duration in epicardium paralleled the restitution of phase 1 amplitude of the action potential in this tissue. The correlation between phase 1 amplitude and action potential duration recorded from a large number of epicardial and endocardial preparations was significant under both steady state and restitution conditions. 4-Aminopyridine, a transient outward current blocker, decreased the time dependence of phase 1 amplitude and concomitantly decreased the time dependence of action potential duration in epicardium. 4-Aminopyridine abbreviated the action potential duration of epicardium at slow stimulation rates but had little effect or prolonged it at fast rates or after premature stimulation. (The availability of a transient outward current is relatively small after premature stimulation.) The data support the hypothesis that the prominent presence of a transient outward current in epicardium, but not endocardium, contributes to the differences in the time and rate dependence of action potential duration and refractoriness in the two tissue types. The results also demonstrate the effect of an outward current to prolong the action potential and the effect of an outward current blocker to abbreviate the action potential.(ABSTRACT TRUNCATED AT 250 WORDS)

Tetrodotoxin-sensitive component in action potential plateau of guinea pig Purkinje fibers: comparison with the papillary muscle

1. The effects of tetrodotoxin on action potentials of isolated guinea pig purkinje fibers were examined and compared the findings with those obtained in the ventricular papillary muscle, by use of conventional microelectrode techniques. 2. Tetrodotoxin (5 x 10(-7)-10(-5) M) decreased the amplitude, overshoot, and maximum upstroke velocity of action potentials of the Purkinje fibers, and shortened the duration of action potential at all levels of repolarization concentration- and stimulus cycle length-dependently. 3. The longer the stimulus cycle length, the greater the shortening by the drug of the action potential duration. 4. In particular, the plateau potential of the Purkinje fibers exposed to tetrodotoxin was remarkably depressed, and which occurred even in case of blockade of K+ conductance, using tetraethylammonium. 5. On the other hand, a high concentration (10(-5) M) of tetrodotoxin did not significantly affect the papillary muscle action potentials. 6. These findings suggest that there is a tetrodotoxin-sensitive component of Na+ current in plateau voltage range of the Purkinje fibers, but little in the papillary muscle, and that the component plays an important role to maintain the plateau of Purkinje fibers action potential.

Mechanism of the use dependence of Ca2+ current in guinea-pig myocytes

1. The mechanism of the use-dependent reduction and facilitation of the calcium current (iCa) in single guinea-pig myocytes described by Fedida, Noble & Spindler (1988) has been examined by varying [Ca2+]o, [Ca2+]i and iCa. 2. Moderate enhancement of [Ca2+]i and [Ca2+]i changes produced by increasing [Ca2+]o reduces iCa and enhances the use-dependent reduction. 3. Intracellular calcium overload, produced by reducing [Na+]o, greatly reduces iCa and almost totally eliminates the use-dependent variations. 4. Use-dependent reduction of iCa is also smaller after substituting external Ba2+ ions for Ca2+ ions. 5. When [Ca2+]i is buffered by intracellular EGTA sufficient to eliminate other [Ca2+]i-dependent processes, such as contraction and Na+-Ca2+ exchange, some use-dependent reduction of iCa remains, although the effect is smaller. Use-dependent facilitation of iCa is more prominent in the presence of internal EGTA. 6. The facilitation of iCa is abolished by Ba2+ replacement of Ca2+ and by the beta-adrenoceptor agonist isoprenaline. This suggests that the facilitation is mediated by Ca2+ entry itself rather than membrane voltage. Facilitation is evident as a delay of current relaxation, even in the presence of internal EGTA.

Use-dependent reduction and facilitation of Ca2+ current in guinea-pig myocytes

1. Action potentials, calcium currents (iCa) and cell contraction have been recorded from single guinea-pig myocytes during periods of stimulation from rest. Voltage clamp was carried out using a single microelectrode. Cell contraction was measured optically. All experiments were performed at 18-22 degrees C. 2. An inverse relationship was observed between cell contraction and action potential duration or iCa. Mixed trains of action potentials and voltage clamp pulses preserved this relationship. Long voltage clamp pulses induced negative 'staircases' of iCa and positive 'staircases' of cell contraction. A facilitation of iCa was observed during repetitive stimulation with clamp pulses of 100 ms duration or less and was accompanied by a decrease in cell contraction. 3. The voltage dependence of inward current staircases was found to depend on Ca2+ entry rather than membrane voltage for long voltage clamp pulses and was not affected by 30 mM-TEA or 50 microM-TTX. Current reduction was greatest at 0 mV (P less than 0.05) when iCa was largest. Changes in cell contraction during pulse trains showed a similar voltage dependence. The time constant of current staircases was only mildly voltage dependent. 4. Interference with normal cellular mechanisms for Ca2+ uptake and release by strontium, 1-5 mM-caffeine and 1 microM-ryanodine increased current staircases and could abolish iCa facilitation with short clamp pulses. 5. Variations in the level of Ca2+-dependent inactivation of iCa can explain many features of the changes in iCa during stimulation after rest. Long clamp pulses (or action potentials) may increase cell Ca2+ loading and inhibit iCa. Short clamp pulses reduce available Ca2+ for cell contraction and this may reflect a lowered myoplasmic Ca2+ level which allows facilitation of iCa.