Mechanism of chloride-dependent release of Ca2+ in the sarcoplasmic reticulum of rabbit skeletal muscle.

We investigated the effect of Cl- on the Ca2+ permeability of rabbit skeletal muscle junctional sarcoplasmic reticulum (SR) using 45Ca2+ fluxes and single channel recordings. In 45Ca2+ efflux experiments, the lumen of the SR was passively loaded with solutions of 150 mM univalent salt containing 5 mM 45Ca2+. Release of 45Ca2+ was measured by rapid filtration in the presence of extravesicular 0.4-0.8 microM free Ca2+ and 150 mM of the same univalent salt loaded into the SR lumen. The rate of release was 5-10 times higher when the univalent salt equilibrated across the SR-contained Cl- (Tris-Cl, choline-Cl, KCl) instead of an organic anion or other halides (gluconate-, methanesulfonate-, acetate-, HEPES-, Br-, I-). Cations (K+, Tris+) could be interchanged without a significant effect on the release rate. To determine whether Cl- stimulated ryanodine receptors, we measured the stimulation of release by ATP (5 mM total) and caffeine (20 mM total) and the inhibition by Mg2+ (0.8 mM estimated free) in Cl(-)-free and Cl(-)-containing solutions. The effects of ATP, caffeine, and Mg2+ were the largest in K-gluconate and Tris-gluconate, intermediate in KCl, and notably poor or absent in choline-Cl and Tris-Cl. Procaine (10 mM) inhibited the caffeine-stimulated release measured in K-gluconate, whereas the Cl- channel blocker clofibric acid (10 mM) but not procaine inhibited the caffeine-insensitive release measured in choline-Cl. Ruthenium red (20 microM) inhibited release in all solutions. In SR fused to planar bilayers we identified a nonselective Cl- channel (PCl: PTris: PCa = 1:0.5:0.3) blocked by ruthenium red and clofibric acid but not by procaine. These conductive and pharmacological properties suggested the channel was likely to mediate Cl(-)-dependent SR Ca2+ release. The absence of a contribution of ryanodine receptors to the Cl(-)-dependent release were indicated by the lack of an effect of Cl- on the open probability of this channel, a complete block by procaine, and a stimulation rather than inhibition by clofibric acid. A plug model of Cl(-)-dependent release, whereby Cl- removed the inhibition of the nonselective channel by large anions, was formulated under the assumption that nonselective channels and ryanodine receptor channels operated separately from each other in the terminal cisternae. The remarkably large contribution of Cl- to the SR Ca2+ permeability suggested that nonselective Cl- channels may control the Ca2+ permeability of the SR in the resting muscle cell.