Influence of permeant ions on voltage sensor function in the Kv2.1 potassium channel

Abstract

We previously demonstrated that the outer vestibule of activated Kv2.1 potassium channels can be in one of two conformations, and that K(+) occupancy of a specific selectivity filter site determines which conformation the outer vestibule is in. These different outer vestibule conformations result in different sensitivities to internal and external TEA, different inactivation rates, and different macroscopic conductances. The [K(+)]-dependent switch in outer vestibule conformation is also associated with a change in rate of channel activation. In this paper, we examined the mechanism by which changes in [K(+)] modulate the rate of channel activation. Elevation of symmetrical [K(+)] or [Rb(+)] from 0 to 3 mM doubled the rate of on-gating charge movement (Q(on)), measured at 0 mV. Cs(+) produced an identical effect, but required 40-fold higher concentrations. All three permeant ions occupied the selectivity filter over the 0.03-3 mM range, so simple occupancy of the selectivity filter was not sufficient to produce the change in Q(on). However, for each of these permeant ions, the speeding of Q(on) occurred with the same concentration dependence as the switch between outer vestibule conformations. Neutralization of an amino acid (K356) in the outer vestibule, which abolishes the modulation of channel pharmacology and ionic currents by the K(+)-dependent reorientation of the outer vestibule, also abolished the K(+)-dependence of Q(on). Together, the data indicate that the K(+)-dependent reorientation in the outer vestibule was responsible for the change in Q(on). Moreover, similar [K(+)]-dependence and effects of mutagenesis indicate that the K(+)-dependent change in rate of Q(on) can account for the modulation of ionic current activation rate. Simple kinetic analysis suggested that K(+) reduced an energy barrier for voltage sensor movement. These results provide strong evidence for a direct functional interaction, which is modulated by permeant ions acting at the selectivity filter, between the outer vestibule of the Kv2.1 potassium channel and the voltage sensor.

Figure 1.

K⁺-dependent increase in current magnitude and activation rate. (A) Two superimposed traces recorded consecutively from a single cell illustrate currents activated in the presence of 0 mM external K⁺ (dashed line) and 10 mM external K⁺ (solid line). (B) Traces in A were normalized to the peak current magnitude to illustrate the speeding of activation in higher [K⁺]. (C) Plot of activation time constants, measured from single exponential fits of the rising phase of the current. Each bar represents the mean ± SEM from eight cells (0 and 10 K⁺ values were obtained in pairs from the same cells).

Figure 2.

Speeding of gating charge movement by K⁺. (A) Gating currents recorded from two different cells, one recorded in the absence of K⁺ (or any other permeant ion; 0 K) and one recorded with 3 mM internal and external K⁺ (3 K). Superimposed on each current trace is the single exponential fit used to obtain the time constant of decay. (B) The area under the currents in A was measured to obtain a plot of the cumulative rate of charge movement. These were then normalized to the maximum charge movement to give a plot of the fractional Q_on. Time constants, calculated from single exponential fits, were nearly identical for the decay phase of the currents in A and the Q_on curves in B (4.86 and 4.85 ms [0 K], 2.36 and 2.37 ms [3 K]).

Figure 3.

Concentration-dependent speeding of gating charge movement by different permeant ions. (A) Pairs of traces illustrate gating currents recorded at 0 mV in the absence of permeant ions and the presence of symmetrical 3 mM K⁺, 3 mM Rb⁺, 3 mM Cs⁺, and 100 mM Cs⁺. Raw gating current magnitudes ranged from 40 pA (3 Cs) to 270 pA (100 Cs). (B) Concentration-response curves for speeding of Q_on by the 3 different permeant ions. Data were fit by Eq. 1 (materials and methods). Calculated EC₅₀ values for K⁺, Rb⁺, and Cs⁺ were: 0.86 ± 0.01 mM (n = 5–11), 1.13 ± 0.06 mM (n = 3–6) and 41.5 ± 0.05 (n = 3–6). The calculated minimum time constant (saturation of the concentration-response curve) for Cs⁺ data was 2.12 ms. Calculated slope values were 1.72, 1.74, and 1.51. Vertical dashed lines are drawn at the calculated EC₅₀ values for K⁺ and Cs⁺.

Figure 4.

Interaction of permeant ions with the selectivity filter. (A) Pairs of consecutively recorded traces illustrate inward currents carried by 130 mM external Na⁺ (Na) or 130 mM Na⁺ plus 1 mM K⁺, Rb⁺, or Cs⁺. Inward Na⁺ current magnitude was 2,400, 2,500, and 1,800 pA in the three pairs of records shown. (B) Concentration-dependent block of inward current by K⁺, Rb⁺, or Cs⁺, applied externally. Data points represent the mean ± SEM of 3–5 cells. (C) Pairs of consecutively recorded traces. The control inward current was recorded in the presence of 130 mM Na⁺ plus 1 mM K⁺, Rb⁺, or Cs⁺. The second trace illustrates currents evoked following equimolar replacement of 30 mM external NMG+ with TEA. Inward current magnitude in the absence of TEA was 420, 900, and 3,200 pA in the three pairs of records shown. (D) Block of inward current by 30 mM TEA in the presence of the indicated concentration of external K⁺, Rb⁺, or Cs⁺. Data points represent the mean ± SEM of 3–6 cells.

Figure 5.

Permeant ion dependence of TEA sensitivity. Ion concentrations on the abscissa represent the concentration of both internal and external permeant ion. Currents were evoked by depolarization to +40 mV, and then TEA was applied to block the current. Changes in block represent changes in TEA efficacy, not potency (see Immke and Korn, 2000). 3 mM TEA was used to block K⁺ and Rb⁺ currents; 100 mM TEA was used to block Cs⁺ currents. Data points represent the mean ± SEM of 3–6 cells, from which a single best fit was calculated using Eq. 1. The calculated maximum block by 3 mM TEA was 32.8 ± 0.3% for K⁺ currents and 33.3 ± 1.4% for Rb⁺ currents. The calculated maximum block by 100 mM TEA in the presence of Cs⁺ was 90.7 ± 1.0%. Calculated EC₅₀ values for K⁺, Rb⁺, and Cs⁺ were: 1.6 ± 0.1 mM, 1.9 ± 0.3 mM, and 44.8 ± 0.9 mM. Vertical dashed lines are drawn at the calculated EC₅₀ values for K⁺ and Cs⁺ concentration-response curves.

Figure 6.

Comparison of Q-V curves for wild-type Kv2.1 and the mutant, Kv2.1 K356G K382V. (A and B) Superimposed (normalized) gating currents from two different cells, one that contained Kv2.1 channels and one that contained Kv2.1 K356G K382V. Panel A illustrates currents evoked by depolarization to +50 mV, and panel B illustrates currents evoked by depolarization to −10 mV. (C) Complete Q-V curves for the two channels, recorded in the absence of permeant ions. Normalized charge, plotted on the ordinate, was calculated from the integrated gating currents and fit by the Boltzmann equation described in materials and methods. Data points represent the mean ± SEM of 3–4 cells at each potential. The calculated V_1/2 values for Kv2.1 and the mutant channel were 15.1 ± 1.5 mV and 12.8 ± 0.8 mV. Slope values were 8.4 and 8.5, respectively.

Figure 7.

K⁺-dependent speeding of gating charge movement is abolished by the K356G mutation. (A) Two superimposed (normalized) currents recorded from two different cells that contained the mutant channel Kv2.1 K356G K382V. One current was recorded in the absence of permeant ions, one in the presence of symmetrical 3 mM K⁺. Time constants for the current decay in each of these two conditions were: 4.80 ± 0.12 ms (n = 5; 0 K) and 4.39 ± 0.11 ms (n = 4; 3 K). (B) Plot of gating current decay time constant in the presence of different symmetrical [K⁺] in wild type Kv2.1 and Kv2.1 K356G K382V. (C) Normalized currents from two cells that contained the mutant channel Kv2.1 K382V. Time constants for the current decay in each of these two conditions were: 4.59 ± 0.06 ms (n = 4; 0 K) and 2.75 ± 0.08 ms (n = 4; 3 K). (D) Normalized currents from two cells that contained the mutant channel Kv2.1 K356G. Time constants for the current decay in each of these two conditions were: 4.45 ± 0.18 ms (n = 4; 0 K) and 4.01 ± 0.11 ms (n = 6; 3 K).

Figure 8.

Restoration of the K⁺-dependent effect on gating current by chemical modification of a cysteine at position 356. (A, top) Normalized gating currents from two cells, one recorded in the absence of permeant ions and one recorded in the presence of symmetrical 3 mM K⁺, that contained the mutant channel, Kv2.1 K356C. (A, bottom panel) Similar to A, except that cells were preincubated with 2 mM MTSET for 5 min (B) Plot of decay time constants for currents recorded under the four conditions in A. Bars represent mean ± SEM for the number of cells shown in parentheses. The horizontal dashed lines are drawn at the time constant values obtained at 0 and 3 mM symmetrical K⁺ with wild-type Kv2.1 channels.

Figure 9.

K⁺ does not speed gating charge movement by shifting voltage-dependence: use of Kv2.1 I379C. (A) Four superimposed (normalized) traces are shown from four different cells containing the mutant channel Kv2.1 I379C. Two traces were recorded in the absence of K⁺ (0 K) and two in the presence of symmetrical 3 mM K⁺ (3 K). One cell in each of these conditions was recorded after application of 100 μM MTSET to cells for 2 min. In these experiments, MTSET was applied during recording, as in panel C inset. (B) Decay time constants under each of the four conditions in A, in addition to that calculated from currents recorded in the presence of symmetrical 10 mM K⁺. Note that at 3 mM K⁺, the speeding of gating charge movement had saturated. Bars represent mean ± SEM of number of cells shown in parentheses. Horizontal lines illustrate decay time constants obtained from Kv2.1 at 0 and 3 mM K⁺. (C) Family of gating currents from a single cell, obtained by depolarization to potentials between −80 and +60 mV after preincubation with 100 μM MTSET for 2 min. (Inset) Illustration of block of ionic current by application of 100 μM MTSET, from a cell recorded in symmetrical 3 mM K⁺. (D) Q-V curve from Kv2.1 I379C under two conditions: in the absence of K⁺ (no MTSET treatment) and in the presence of symmetrical 3 mM K⁺ after pretreatment with MTSET. Calculated V_1/2 values were −16.6 ± 2.7 mV (n = 3) and −16.4 ± 1.9 mV (n = 3). Slope values were 13.4 and 14.9.

Figure 10.

[K⁺]-dependent change in Q_on kinetics as a function of voltage. Gating currents were recorded in Kv2.1 I379C under two conditions: 0 internal and external K⁺ (filled circles) and 3 mM symmetrical K⁺ plus MTSET (open squares). (A) Decay time constant as a function of voltage. Each data point represents the mean of measurements obtained from 3 to 13 cells. (B) Semilogarithmic plot of the data in A. Linear regression curves were fit for data between −10 and +50 mV. (C and D) α and β as a function of voltage, calculated as described in the text. (E) Ratios of α and β, from the data in C and D.

Figure 11.

[K⁺]-dependent increase in rate of ionic current activation. (A) Superimposed, (normalized) ionic currents from Kv2.1 recorded at +40 mV in the presence of symmetrical 0.3 mM K⁺ and 10 mM K⁺. Peak current magnitudes were 110 pA (0.3K) and 4,400 pA (10K). Due to the large transient, the first 5 ms of the 0.3 mM K⁺ current were blanked. (B) Plots of time constants of ionic current activation and gating current decay at different symmetrical [K⁺]. Data represent mean ± SEM of 4–6 cells.