1. The functional role of reverse Na+-Ca2+ exchange in the activation of contraction of rat ventricular myocytes has been studied. Mechanical activity of single cells, measured as unloaded cell shortening, was recorded simultaneously with membrane current and voltage using a single microelectrode voltage clamp and a video edge detection device. 2. The voltage dependence of contraction was studied by applying trains of depolarizations. At test potentials between + 20 and + 80 mV (under conditions where large outward currents were activated) a plateau on the shortening vs. voltage (S-V) relationship was observed. Significant cell shortening also occurred at test potentials between - 70 and - 40 mV; and these contractions were accompanied by large inward Na+ currents. We have investigated the ionic mechanisms for three components of the S-V relation in rat ventricle: (i) shortening which occurs between - 70 and - 40 mV and is thought to be dependent on the sodium current; (ii) phasic contractions in the voltage range - 40 to + 40 mV where the L-type Ca2+ current is present; (iii) the plateau of the S-V relation at strongly depolarized voltages where reverse Na+-Ca2+ exchange may occur. 3. Experiments in which two independent microelectrode impalements were made in a single myocyte showed that during activation of contraction at test potentials between -70 and -40 mV, and during very large depolarizations (+20 to +80 mV), there were significant deviations of the measured membrane potential from the applied voltages. Activation of cell shortening in these voltage ranges could be eliminated by electronic series resistance compensation, which significantly reduced these voltage errors. Consistent with these findings, when tetrodotoxin (TTX) and 4-aminopyridine (4-AP) were used to block inward Na+ and transient outward K+ currents, respectively, no significant voltage errors were present and a bell-shaped shortening-voltage (S-V) relationship was obtained. 4. When Na+ and K+ currents were blocked, depolarizations from holding potentials of either -80 or -50 mV demonstrated that the threshold for activation of contraction was about -30 mV, and that the voltage dependence of peak shortening was very similar to that of the L-type Ca2+ current (I(Ca, L)). These contractions were suppressed completely by either Cd2+ or ryanodine, showing that activation of cell shortening was due to Ca2+ influx through L-type channels which currents were observed. 5. The bell-shaped S-V relationship remained unchanged when the Ca2+ fluxes due to Na+-Ca2+ exchanger were altered significantly by adjusting the electrochemical gradient for Na+. In some of these experiments, extracellular Na+ was rapidly reduced using Li+ substitution; in others intracellular Na+ was varied between 5 and 15 mm, by increasing [Na+] in the recording microelectrode. 6. Additional experiments were performed to determine whether Na+-Ca2+ exchange operating in reverse mode could trigger SR Ca2+ release. Ca2+ influx due to I(Ca,L) was blocked with CdCl2, active uptake of Ca2+ by the SR was reduced using caffeine, and Ni2+ was used as an inhibitor for Na+-Ca2+ exchange. Under these conditions a slow Ni2+-sensitive component of Ca2+ entry at membrane potentials positive to -20 mV was observed during 1-2 s depolarizations. The resulting changes in cell length were very small and slow; rapid phasic (twitch) contractions were never elicited by these depolarizations. Thus, reverse Na+-Ca2+ exchange cannot bring enough Ca2+ into the cytosol to activate a regenerative Ca2+ release from intracellular stores. 7. These results demonstrate that activation of contraction in rat ventricular myocytes is dependent on SR Ca2+ release which is initiated by Ca2+ influx through L-type Ca2+ channels. Ca2+ influx through either T-type Ca2+ channels or via reverse Na+-Ca2+ exchange cannot trigger normal phasic contractions.
