Modification by protons of frog skeletal muscle KATP channels: effects on ion conduction and nucleotide inhibition.

1. The molecular mechanisms underlying pH regulation of skeletal muscle ATP-sensitive K+ (KATP) channels were studied using the patch clamp technique in the inside-out configuration. Two effects of intracellular protons were studied in detail: the decrease in magnitude of single-channel currents and the increase in open probability (Po) of nucleotide-inhibited channels. 2. The pH dependence of inward unit currents under different ionic conditions was in poor agreement with either a direct block of the pore by protons or an indirect proton-induced conformational change, but was compatible with the protonation of surface charges located near the cytoplasmic entrance of the pore. This latter electrostatic mechanism was modelled using Gouy-Chapman-Stern theory, which predicted the data accurately with a surface charge density of about 0.1 negative elementary charges per square nanometre and a pK (pH value for 50% effect) value for protonation of these charges of 6.25. The same mechanism, i.e. neutralization of negative surface charges by cation binding, could also account for the previously reported reduction of inward unit currents by Mg2+. 3. Intracellular alkalization did not affect Po of the KATP channels. Acidification increased Po. In the presence of 0.1 mM ATP (no Mg2+), the channel activation vs. pH relationship could be fitted with a sigmoid curve with a Hill coefficient slightly above 2 and a pK value of 6. This latter value was dependent on the ATP concentration, decreasing from 6.3 in 30 microM ATP to 5.3 in 1 microM ATP. 4. Conversely, the channel inhibition vs. ATP concentration curve was shifted to the right when the pH was lowered. At pH 7.1, the ATP concentration causing half-maximal inhibition was about 10 microM. At pH 5.4, it was about 400 microM. The Hill coefficient values remained slightly below 2. Similar effects were observed when ADP was used as the inhibitory nucleotide. 5. These results confirm that a reciprocal competitive link exists between proton and nucleotide binding sites. Quantitatively, they are in full agreement with a steady-state model of a KATP channel possessing four identical protonation sites (microscopic pK, 6) allosterically connected to the channel open state and two identical nucleotide sites (microscopic ATP dissociation constant, approximately 30 microM) connected to the closed state.