Voltage‐dependent magnesium block of adenosine‐triphosphate‐sensitive potassium channel in guinea‐pig ventricular cells.

Abstract

1. The adenosine‐5'‐triphosphate (ATP)‐sensitive K+ channel of guinea‐pig ventricular cells was examined in the presence and absence of internal Mg2+ or Na+ using an open cell‐attached configuration of the patch‐clamp technique. 2. Millimolar concentrations of internal Mg2+ ([Mg2+]i) produced marked fluctuations in the outward current, and the amplitude of the open‐channel current was reduced with increasing [Mg2+]i. Millimolar Na+ applied internally also decreased the mean amplitude of the outward current, but the increase in current noise was not obvious. These effects became larger when the membrane potential was shifted to be more positive from the K+ equilibrium potential (EK). At potentials negative to EK the inward current was affected by neither internal Mg2+ nor Na+. 3. The external application of Na+, Mg2+ or Ca2+, however, failed to affect the single‐channel current. 4. After removal of both internal Mg2+ and Na+, the mean open‐channel current‐voltage relationship became virtually linear. Referring to these unblocked values, relative amplitudes were determined at different levels of [Mg2+]i or [Na+]i. The dose‐response relations gave a Hill coefficient of approximately 1 for Mg2+ block and approximately 2 for Na+ block. The half‐maximum concentrations (Kh) for both Mg2+ and Na+ block were shifted to lower values with increasing positive potentials. 5. The power‐density spectrum of the open‐channel current noise induced by internal Mg2+ showed a Lorentzian function with a corner frequency above 1 kHz, suggesting that the current noise is due to rapid fluctuations of open‐channel current between blocked and unblocked states. The corner frequencies gave Mg2+ block and unblock rate constants which were of the order of 10(7) M‐1 s‐1 and 10(4) s‐1, respectively. 6. With increasing external K+ concentration ([K+]o) from 0 to 140 mM the current fluctuations became less prominent, and Kh for Mg2+ block was shifted to higher values. Raising [K+]o enhanced the unblock rate derived from the noise analysis while the block rate was not significantly altered. 7. The above findings could be explained by assuming a binding site for one Mg2+ or two Na+ located 30‐35% of the electrical drop across the membrane from the inner mouth of the channel, thereby resulting in the ionic block of K+ passage. An apparent inward rectification observed in the single‐channel current‐voltage relation is attributable to the blockade of the channel by intracellular Mg2+ and/or Na+.