1. The effect of myoplasmic Mg2+ on Ca2+ release was examined in mechanically skinned skeletal muscle fibres, in which the normal voltage-sensor control of Ca2+ release is preserved. The voltage sensors could be activated by depolarizing the transverse tubular (T-) system by lowering the [K+] in the bathing solution. 2. Fibres spontaneously contracted when the free [Mg2+] was decreased from 1 to 0.05 mM, with no depolarization or change of total ATP, [Ca2+] or pH (pCa 6.7, 50-mu-M-EGTA). After such a 'low-Mg2+ response' the sarcoplasmic reticulum (SR) was depleted of Ca2+ and neither depolarization nor caffeine (2 mM) could induce a response, unless the [Mg2+] was raised and the SR reloaded with Ca2+. Exposure to 0.05 mM-Mg2+ at low [Ca2+] (2 mM-free EGTA, pCa > 8.7) also induced Ca2+ release and depleted the SR. 3. The response to low [Mg2+] was unaffected by inactivation of the voltage sensors, but was completely blocked by 2-mu-M-Ruthenium Red indicating that it involved Ca2+ efflux through the normal Ca2+ release channels. 4. In the absence of ATP (and creatine phosphate), complete removal of Mg2+ (i.e. no added Mg2+ with 1 mM-EDTA) did not induce Ca2+ release. Depolarization in the absence of Mg2+ and ATP also did not induce Ca2+ release. 5. Depolarization in 10 mM-Mg2+ (pCa 6.7, 50-mu-M-EGTA, 8 mM-total ATP) did not produce any response. In the presence of 1 mM-EGTA to chelate most of the released Ca2+, depolarizations in 10 mM-Mg2+ did not noticeably deplete the SR of Ca2+, whereas a single depolarization in 1 mM-Mg2+ (and 1 mM-EGTA) resulted in marked depletion. Depolarization in the presence of D600 and 10 mM-Mg2+ produced use-dependent 'paralysis', indicating that depolarization in 10 mM-Mg2+ did indeed activate the voltage sensors. 6. Depolarization in the presence of 10 mM-Mg2+ and 25-mu-M-ryanodine neither interfered with the normal voltage control of Ca2+ release nor caused depletion of the Ca2+ in the SR even after returning to 1 mM-Mg2+ for 1 min, indicating that few if any of the release channels had been opened by the depolarization. 7. In solutions with 10 mM-Mg2+ and elevated [Ca2+] (1.5-mu-M-Ca2+, 50-mu-M-EGTA), depolarization produced little if any response, but depolarization with simultaneous application of caffeine (2 mM) produced a large response, indicating that the channels were not merely 'blocked' by the elevated [Mg2+], but that the depolarization alone could not substantially activated the channels. 8. These results show that the release channels are normally inhibited in the presence of 1 mM-Mg2+, and that activating the voltage sensors (by depolarization) can overcome this inhibition and open the release channels in 1 mM-Mg2+ but not in 10 mM-Mg2+. We suggest that activation of the voltage sensors may lower the affinity of the release channels for Mg2+, thereby removing the resting inhibition and activating the channels. This proposal helps to reconcile and unify apparently contradictory mechanisms of voltage-sensor control of Ca2+ release and Ca2+-induced Ca2+ release.
