Rectification and Rapid Activation at Low Ca²⁺ of Ca²⁺-Activated, Voltage-Dependent BK Currents: Consequences of Rapid Inactivation by a Novel β Subunit

Distribution of β3 message in various tissues.A, Northern blots show membranes containing human message probed with radiolabeled human β3 sequence. Tissues are as follows: 1, heart; 2, brain;3, placenta; 4, lung; 5, liver; 6, skeletal muscle; 7, kidney;8, pancreas; 9, adrenal medulla;10, thyroid; 11, adrenal cortex;12, testis; 13, thymus;14, small intestine; 15, stomach. The minimal predicted transcript size for the β3 variant described here would be expected to be ∼1 kb.

Coexpression of the β3 subunit with theSlo α subunit produces rapidly inactivating, Ca²⁺- and voltage-dependent current.A, Traces show currents obtained from an inside-out patch from a Xenopus oocyte injected with cRNA encoding both human β3 and mouse α subunits. Channels were activated by voltage steps from −140 to +180 mV with 0, 1, 4, 10, and 300 μm Ca²⁺. The voltage protocol is shown at the top. B, Faster time base records of the currents in A are shown for both activation and inactivation and for deactivation. Note that the largest tail current amplitude at −120 mV greatly exceeds the steady-state current level during even the most positive activation steps, indicating that there must be extensive channel unblocking before the peak of the tail current. Also, note that, at 1 and 4 μmCa²⁺, the maximal tail current at −120 mV is larger than the maximal peak current activated during the voltage step to +180 mV. This suggests that, at lower Ca²⁺, many channels become blocked during the rising phase of outward current.

Voltage and Ca²⁺ dependence of the inactivation time constant of β3 currents. A, Traces on the left show normalized currents (from same patch as Fig. 3) activated with 300 μmCa²⁺ for +60, +100, +140, and +180 mV. In each case, the best fit of a single exponential function to the current inactivation time course is plotted over the trace, yielding the time constant of inactivation (τ_i). τ_iwas 1.26, 0.77, 0.62, and 0.57 msec for +60, +100, +140, and +180 mV, respectively. Traces on the right show normalized currents (also from Fig. 3) activated at +140 mV for 1, 4, 10, and 300 μm Ca²⁺. Fitted single exponential functions are also plotted. τ_i was 1.49, 0.74, 0.65, and 0.62 msec for 1, 4, 10, and 300 μm, respectively.B, τ_i is plotted as a function of command potential for four different [Ca²⁺]. For each Ca²⁺, there is a small increase in apparent inactivation rate with more depolarized command potential. Each point is the mean ± SEM with four patches for 1 μmCa²⁺, 10 patches at 10 μm, nine patches at 300 μm Ca²⁺, and five patches for 1 mm Ca²⁺.C, τ_i is replotted as a function of [Ca²⁺]. At any voltage, there is little indication of any Ca²⁺ dependence to τ_i at Ca²⁺ of 10 μm and higher.

Comparison of currents resulting from α alone, α + β3, and α + β1. A, Tracesshow currents resulting from expression of α alone activated by the indicated voltage protocol with 10, 60, and 300 μmCa²⁺. B, Traces show α + β3 currents under the same conditions as in A. C, Traces show currents resulting from α + β1 expression under the same conditions as in A andB, except that the potential both before and after the activation step was −180 mV. The α + β3 combination results in currents that, at a given Ca²⁺, are activated at more negative potentials than for α subunits alone but somewhat more positive than for α + β1 subunits. This can be most clearly seen in the tail currents. At 60 μm, the step to −120 mV results in only slight current activation for the β3 construct but produces almost 50% activation for the β1 construct.

The voltage dependence of activation and steady-state properties of inactivation revealed in conductance–voltage curves. A, Examples ofG–V curves obtained from tail current amplitude (open triangles), from the peak current (inverted filled triangles) measured during the command step, and from the steady-state current at the end of the voltage step (filled triangles) are shown for one patch bathed with 300 μm Ca²⁺. For tail currents, each point corresponds to the conductance measured 110 μsec after the nominal imposition of the repolarizing voltage step to −120 mV from a given command potential. Reduction of conductance at the most positive potentials probably reflects some slow cumulative block from contaminating ions. B, A repolarizing step to −160 mV was used to compare tail current amplitude and time course either from just after the peak of outward current during a step to +160 mV or after development of the steady-state current level at +160 mV. Despite the fact that the current near the peak is over twofold larger than the steady-state current, the tail current amplitude in each case is essentially identical. τ_d after repolarization near the peak outward current was 0.70 msec, whereas after repolarization from the steady-state current level it as 0.97 msec. C, Tail current conductances (from the same patch as in A) were determined for 0 (filled circles), 0.5 (open circles), 1 (filled diamonds), 4 (open diamonds), 10 (filled triangles), and 300 (open triangles) μmCa²⁺ . The solid lines are single Boltzmann fits (Eq. 1) to the G–V curves. Values were as follows: for 0 μm,V_0.5 = 113.3 mV, k= 15.3 mV; for 0.5 μm,V_0.5 = 67.1 mV, k = 15.97; for 1 μm, V_0.5 = 40.4 mV, k = 15.4 mV; for 4 μm,V_0.5 = 16.2 mV, k = 15.9 mV; for 10 μm, V_0.5= −32.2 mV, k = 16.08; for 300 μm,V_0.5 = −70.5, k = 17.47 mV. D, Steady-state conductances were determined from the average current level at the end of voltage steps to a given activation potential and plotted as a function of command voltage.Solid lines are fits of the double Boltzmann function given by Equation 2. Values for V_0.5 for activation and k for activation were constrained to those obtained in C. E, Peak conductances were calculated from peak currents at each activation potential using a 0 mV reversal potential and plotted as a function of command potential. The lines simply connect the values. F, Voltages at which conductance is half-activated are plotted as a function of [Ca²⁺] for α alone, α + β1, α + hβ3, and α + β2(Δ33) in which the β2 inactivation has been deleted. Error bars show SD. For α + β3, points at 0, 10, and 300 μm correspond to 17 patches, points at 1, 4, 1000, and 5000 μm correspond to 6–10 patches, and points at 0.5 μm corresponds to three patches. For α alone, each point corresponds to four patches, for α + β1, six patches, and for α + β2(Δ33), at least five patches for each point. For graphical purposes, points in nominally 0 Ca²⁺were plotted at 0.05 μm.

Channel openings resulting from coexpression of α + β3 and exhibiting unusual gating behavior.Traces in each column show channel openings in an inside-out patch bathed with 0, 10, 60, or 300 μm Ca²⁺, respectively, fromleft to right. Channel openings were activated by steps to +100 mV (with the voltage protocol shown on thetop of the first column). Four consecutive sweeps are shown below each condition. Thedotted line indicates a 25 pA current level (250 pS at this voltage). The records were filtered at 10 kHz, and currents were sampled at 100 kHz. Traces on the bottomof each column show ensemble averages for each condition shown above. In each case, at least 60 sweeps were included in each average. Averages exhibit the rapid but incompletely inactivating current seen for the macroscopic currents, despite the fact that no fully resolved channel openings are seen. Single exponential fits to the inactivation time course of the averaged currents yielded values of 0.99, 1.37, and 1.18 msec for 10, 60, and 300 μm Ca²⁺, respectively, comparable with the inactivation time constants of macroscopic currents. Calibration: 30 pA applies to the individual current sweeps, whereas 15 pA applies to the averaged currents.

β3 currents exhibit sensitivity to TEA characteristic of other BK currents. A,Traces show currents activated by the indicated voltage-protocol (activation steps from −100 to +100 mV) from an outside-out patch from an oocyte expressing the Slo α subunit alone with 10 μm Ca²⁺ in the recording pipette. Left panel shows currents in control saline, middle panels show the effects of 1 and 30 mm TEA applied to the extracellular face of the membrane, and the right panel shows currents after removal of TEA.B, Traces show currents activated with the same protocol used in A from an outside-out patch expressing α + β3 subunits. C, The fractional block of current elicited at +100 mV is plotted as a function of extracellular TEA concentration for both α alone (open circles) and α + β3 (filled circles). Fractional block was normalized to the amount of block produced by 30 mm TEA. The IC₅₀ values for block were 0.43 ± 0.05 mm (n= 3) for α with a Hill coefficient of 0.82, and 0.59 ± 0.30 mm (n = 4) for block of α+ hβ3 with a Hill coefficient of 0.94.D, Estimates of the IC₅₀ for block by TEA are plotted as a function of voltage. Fractional block in each case was measured from the steady-state currents at +20 through +100 mV.Solid lines are fits of the functionK(V) = K(0) * exp(−zFV/RT), whereK(0) indicates the IC₅₀ for block by TEA at 0 mV, F, R, and T have their usual meanings, and z is the fitted value for fractional charge moved during the blocking reaction. For α,K(0) = 0.18 ± 0.01 mm withz = 0.28 ± 0.025, whereas for α + β3,K(0) = 0.22 ± 0.04 mm, withz = 0.30 ± 0.076.

The β3 N terminus, but not the C terminus, is responsible for inactivation of the α + β3 currents.A, Thirty-five amino acids from the C terminus of the β3 subunit (construct D3) were removed, and the resulting construct was coexpressed with α subunits. Currents retain the typical inactivation behavior of the intact β3 subunit. Currents were activated with the indicated voltage protocol with 10 μmcytosolic Ca²⁺. B, Twenty-one amino acids were removed from the N terminus of the β3 subunit (construct D4), resulting in loss of the inactivation behavior (10 μm Ca²⁺). C, Twenty-one amino acids from the β3 N terminus were used to replace 33 amino acids at the N terminus of the β2 subunit (construct D20). The D20 construct exhibits the rapid, incomplete inactivation of the β3 subunit but the more negative activation range of the β2 subunit. Currents were activated with 10 μmCa²⁺. D, The Gmhβ3 alternative splice variant (Riazi et al., 1999) was expressed, and currents were activated with 10 μm Ca²⁺. Note the different time base for currents in D. The time constant of inactivation for this construct at 10 μm and +100 mV was ∼50 msec, although the time course of decay is somewhat better described with two exponential components.

Cytosolic application of TEA does not compete with the native β3 inactivation process. A,Traces show β3 currents activated by the indicated voltage step with 300 μm cytosolic Ca²⁺ and 0, 30, 60, or 100 mm TEA. A trace was also recorded in 0 Ca²⁺ with 100 mm added TEA. TEA results in a concentration-dependent reduction of peak and steady-state outward current but no effect on the time constant of inactivation. B, The fold increase of τ_i by TEA and fold decrease in peak current amplitude is plotted as a function of TEA. For a simple competition between TEA and the inactivation mechanism, both relationships should increase with [TEA] in an identical manner. Larger symbols are means with SDs for estimates from five patches. Thesmaller symbols represent individual determinations from these five patches.