1. Cytosolic free calcium ion concentration ([Ca2+](i)) and whole-cell
L-type Ca2+ channel currents were measured during excitation-contract
ion (E-C) coupling in single voltage-clamped rat cardiac ventricular c
ells. The measurements were used to compute the total cellular efflux
of calcium ions through sarcoplasmic reticulum (SR) Ca2+ release chann
els (F-SR,F-rel) and the influx of Ca2+ via L-type Ca2+ channels (F-Ic
a). 2. F-SR,F-rel was elicited by depolarizing voltage-clamp pulses 20
0 ms in duration to membrane potentials from -30 to +80 mV. Over this
range, peak F-SR,F-rel had a bell-shaped dependence on clamp pulse pot
ential. In all cells, the 'gain' of the system, measured as the ratio,
F-SR,F-rel(max)/F-Ica(max), declined from about 16, at 0 mV, to much
lower values as clamp pulse voltage was made progressively more positi
ve. We named this phenomenon of change in gain as a function of membra
ne potential, 'variable gain'. At clamp pulse potentials in the range
-30 to 0 mV, the gain differed from cell to cell, being constant at ab
out 16 in some cells, but decreasing from high values (approximate to
65) at -20 mV in others. 3. At clamp pulse potentials at which Ca2+ in
flux (F-ICa) was maintained, F-SR,F-rel also had a small maintained co
mponent. When macroscopic Ca2+ influx was brief (1-2 ms, during 'tails
' of F-Ica),F-SR,F-rel rose rapidly to a peak after repolarization and
then declined with a half-time of about 9 ms (typically). 4. The risi
ng phase of[Ca2+](1) transients could be interrupted by stopping Ca2influx rapidly (by voltage clamp). We therefore termed this phenomenon
'interrupted SR Ca2+ release'. 5. 'Variable gain', high gain in a sta
ble system, and 'interrupted SR Ca2+ release' are not consistent with
'common pool' theories of E-C coupling. Therefore the results are disc
ussed in terms of a theory in which SR Ca2+ release channels are contr
olled locally by Ca2+ influx through co-associated L-type Ca2+ channel
s and by Ca2+ released from neighbouring SR release channels.