Triple-barrel structure of inwardly rectifying K+ channels revealed by Cs+ and Rb+ block in guinea-pig heart cells.

1. The hypothesis that the inwardly rectifying K+ channel consists of a triple-barrel structure was investigated. Inward currents were recorded under the blocking effects of external Cs+ or Rb+ in the cell-attached configuration of the patch-clamp technique using single ventricular cells enzymatically isolated from guinea-pig hearts. 2. Cs+ (10-100 microM) or Rb+ (20-100 microM) added to the 150 mM-K+ pipette solution induced rapid open-blocked transitions in the inward open-channel currents. In about 20% of experiments the inward current showed two intermediate current levels equally spaced between the unit amplitude and the zero-conductance level. The current fluctuated between these four levels. In the remaining experiments no obvious sublevels were observed except spontaneous ones, whose amplitudes were not always equal to one-third or two-thirds of the unit amplitude. 3. In experiments showing sublevels, the probability that the open-channel current stayed at each level was measured at various concentrations of blockers and membrane potentials. In both Cs+ and Rb+ block, the distribution of the current levels showed reasonable agreement with the binomial theorem. This finding suggests that the inwardly rectifying K+ channel is composed of three equally conductive subunits and each subunit is independently blocked by Cs+ or Rb+. 4. The dwell-time histogram in each substate was well fitted with a single-exponential function. On the assumption of the binomial model, the blocking (mu) and unblocking (lambda) rate for Cs+ and Rb+ were calculated. The value of mu was linearly proportional to the concentration of the blocking ion at a given membrane potential and increased with hyperpolarization (e-fold increase with a change of -43.5 mV in the Cs+ block). lambda was almost independent of the concentration of the blocking ion and less dependent on the membrane potential than mu. 5. The open and blocked times were calculated in experiments showing no clear sublevels. The mean open time was almost equal to the mean dwell time at the full open level in experiments showing sublevels under the same conditions. On the other hand, the mean blocked time was about two or three times longer than the mean dwell time at the zero-conductance level measured in experiments with sublevels. These results may suggest that the instant one of the three subunits is plugged by blocking ions, the remaining two subunits are closed by unknown mechanisms. 6. Our results support the hypothesis that the cardiac inwardly rectifying K+ channel is composed of three equally conductive subunits.