1. Single Purkinje cells, enzymatically isolated from rabbit ventricle, were studied under whole-cell voltage clamp and internally perfused with the fluorescent Ca2+ indicator, indo-1 (100 muM). 2. Fast [Ca2+]i transients were elicited by brief depolarizations from a holding voltage of -45 mV and by repolarization from very positive potentials. The peak [Ca2+]i-voltage relation was bell-shaped with a peak around +10 mV. 3. [Ca2+]i transients were completely blocked by the Ca2+ channel antagonist, nisoldipine (10 muM) and were very small when Ca2+ release from the sarcoplasmic reticulum (SR) was prevented by superfusion of cells by caffeine (1 mm) or ryanodine (10 muM). A fast application of caffeine induced a transient increase in [Ca2+]i. These results suggest [Ca2+]i transients are due to Ca2+-induced Ca2+ release from the SR. 4. Rate of decline of the [Ca2+]i transient was voltage dependent, suggesting contribution of the Na+-Ca2+ exchanger to Ca2+ efflux. At very positive potentials (> +60 mV), Ca2+ influx through the Na+-Ca2+ exchanger could be observed. 5. A transient outward current was observed at potentials positive to +10 mV, but only if depolarizing pulses were accompanied by a [Ca2+]i transient. 6. When the amplitude of the [Ca2+]i transient was changed by (1) changes in [Ca2+]o, (2) changes in frequency of depolarization or (3) conditioning prepulses, the amplitude of the outward current changed in the same direction. This suggests activation of the current is dependent on and graded by [Ca2+]i. 7. The outward current was observed in K+-free solutions, in the presence of Cs+ and TEA+, and was not blocked by 4-aminopyridine (10 mm). In contrast, DIDS (100 muM) decreased the outward current by 70 +/- 20 % (mean +/- S. D., n = 9), without affecting [Ca2+]i. 8. When external Cl- was lowered, the amplitude of the outward current decreased; when internal Cl- was replaced by aspartate, it became apparent at more negative potentials. These interventions strongly suggest the current was carried by Cl-; it can therefore be referred to as a [Ca2+]i-activated Cl-current or I(Cl(Ca)). 9. When I(Cl(Ca)) was maximally activated during a conditioning step, steps to negative potentials revealed inward currents through I(Cl(Ca)) (in symmetrical Cl- solutions). The fully activated I-V relation was linear. 10. I(Cl(Ca)) could be activated at membrane potentials between -80 and +80 mV by a fast application of caffeine (10 mm), inducing Ca2+ release from the SR, demonstrating that I(Cl(Ca)) does not require membrane depolarization or Ca2+ influx through the Ca2+ channel for its activation. However, at negative potentials the amplitude of I(Cl(Ca)) decreased. 11. The time course of I(Cl(Ca)) did not follow the time course of the [Ca2+]i transient. I(Cl(Ca)) declined early, before [Ca2+]i reached its peak. Study of inward tail currents, and of I(Cl(Ca)) activated by caffeine-induced [Ca2+]i transients, confirmed early decline. 12. These observations cannot be reconciled with a simple ligand-operated channel. The transient nature of I(Cl(Ca)) might be due to an inactivation process, or, alternatively, the differences in time course between I(Cl(Ca)) and [Ca2+]i could reflect the existence of important gradients for Ca2+ between the cytoplasm and the subsarcolemmal space.
