1. The Ca2+-dependent kinetics of large-conductance Ca2+-activated K+ channels from cultured rat skeletal muscle were studied with the patch clamp technique. Data were collected in the absence of Na+ and Mg2+, which can alter the kinetics. About 2 x 10(5) open and shut intervals were analysed from each of five different excised membrane patches containing a single active channel. Analysis was restricted to activity in the normal mode, which includes 96% of the intervals. 2. The open probability (P(open)) and dwell-time distributions of open and shut intervals were obtained at three to four different [Ca2+]i for each of the channels. P(open) data were also obtained from some multichannel patches. 3. Increasing [Ca2+]i increased P(open). At a pH of 7.0 the Hill coefficient was 3.7 +/- 0.8 (range of 3.0-5.0) and a P(open) of 0.5 occurred at 14 +/- 7-mu-M [Ca2+]i (K0.5) for data obtained at +30 mV (n = 6). At a pH of 7.2 the Hill coefficient was 3.0 +/- 0.5 (range of 2.2-3.7) and K0.5 was 9 +/- 6-mu-M-Ca2+ (n = 7). The large standard deviations for K0.5 reflect the observation that fourfold differences in K0.5 could be observed for different channels studied under the same experimental conditions. 4. Hill coefficients that can be greater than 3 suggest that the channel may bind four or more Ca2+ to become fully activated. The binding of four Ca2+ before opening would require a minimum of five shut states. This estimate of the minimum number of shut states is in general agreement with that obtained from the number of exponential components in the dwell-time distributions of shut intervals. Thus, two different methods give similar estimates of the minimum number of shut states. If the channel can open with different numbers of bound Ca2+, then this could give rise to the three to four open states suggested by the three to four exponential components in the open dwell-time distributions. 5. Kinetic schemes consistent with the Ca2+-dependent kinetics were developed by simultaneously fitting open and shut dwell-time distributions obtained at three to four different [Ca2+]i, using maximum likelihood techniques and corrections for missed events. Such simultaneous fitting can provide an increased ability to define models and rate constants. 6. The simplest gating mechanism that could describe the major features of the Ca2+-dependence of P(open), the dwell-time distributions, and the relationship between the durations of the adjacent intervals was: [GRAPHICS] where C and O represent closed and open states, respectively. In general, the lifetimes of the shut states decrease from C to C3Ca, with C4Ca then increasing, and the lifetimes of the open states increase from O2Ca to O4Ca ([Ca2+]i < 10-20-mu-M). The Ca2+ dependence of the channel arises because increasing the [Ca2+]i drives the activity of the channel from the longer shut states and briefer open states towards the briefer shut states and longer open states. Models similar to the above scheme, but expanded to include six shut and four open states could improve the description of the data. 7. The effective rate constant for Ca2+ binding increased about 6-fold for the binding of the second Ca2+ and then increased another 60-fold for the binding of the third Ca2+, suggesting co-operative interaction among the binding sites. Models with identical and independent Ca2+ bindings sites were inconsistent with the data. 8. The kinetic scheme in paragraph 6 and the expanded versions of this scheme provide working hypotheses that can describe the major features of Ca2+-dependent channel gating over a 400-fold range of P(open).
