|
Figure 1. Activation and inactivation voltage dependence of Kv4.2 wild-type channels. (A) A-type currents in response to depolarizing voltage jumps from −110 mV to different test potentials between −90 and +90 mV in 10-mV increments. Currents were recorded under two-electrode voltage clamp in Xenopus oocytes. (B) Prepulse inactivation of outward currents at +40 mV induced by conditioning prepulses of 20-s duration to potentials between −120 and 0 mV in 5-mV increments. (C) Voltage dependence of steady-state inactivation (filled symbols; n = 9) and voltage dependence of relative peak conductance activation (open symbols; n = 7) reflect the channel availability. The inactivation curve represents a first-order Boltzmann function, the activation curve a fourth-order Boltzmann function.
|
|
Figure 2. Kinetic inactivation parameters of Kv4.2 wild-type channels. (A) Onset of low-voltage inactivation. Outward currents obtained with brief voltage pulses to +40 mV, either directly from −110 mV (Icontrol) or after a conditioning prepulse of different length (between 25 ms and 51.2 s) to −65 mV (Itest). Note the time axis breaks and indicated durations for extremely long prepulses. (B) Relative current amplitudes (Itest/Icontrol) obtained with different prepulse durations at −65 mV (n = 8). A single-exponential fit to the data yields the time constant of onset of low-voltage inactivation (τL) and the fraction of non-inactivated channels (f). Single-exponential fitting curves for the data obtained with other prepulse potentials are shown in gray (see Fig. S2 A for complete datasets). (C) Recovery from low-voltage inactivation. After a voltage pulse from −100 to +40 mV, which elicits the control current, low-voltage inactivation is induced at −55 mV for 20 s. Then, after different amounts of time at −100 mV, the test current amplitude at +40 mV is measured. (D) Relative current amplitudes (test/control) obtained with different interpulse durations expressed as percent recovery from inactivation at −100 mV (n = 7). A single-exponential fit yields the time constant of recovery from low-voltage inactivation (τrec). Single-exponential fitting curves for the data obtained at other recovery potentials are shown in gray (see Fig. S2 B for complete datasets). Onset and recovery curves obtained at −80 mV in B and D, respectively (dark gray), represent equal time constants (see also Fig. S2 C).
|
|
Figure 3. Voltage sensor gate uncoupling model and putative protein domains involved. (A) Conceptual gating scheme illustrating a working model for the closed-state inactivation of Kv4.2 channels (black letters). In each α-subunit, depolarization leads to a transition from “Resting” (R) to “Activated” (A). If all four subunits reach the activated state (AAAA), the channel can open (O; gray letters). However, the transition of a single subunit to an inactivated/desensitized (I/D) state is sufficient to prevent channel opening. Cartoons depict the three states of an individual α-subunit with its six-membrane–spanning domains (S1-S6; positive charge in S4) and cytoplasmic N and C termini. Resting (R), voltage sensor down, S6 gate closed; activated (A), voltage sensor up (vertical green arrow) and S6 gate open permissive (blue arrow pointing to the left); inactivated/desensitized (I/D), voltage sensor still up, but in a more stable conformation than before, and S6 gate closed again but uncoupled from the voltage sensor (green and blue arrow pointing to the right). “Inactivated” refers to the voltage sensor conformation, and “Desensitized” refers to the state of the uncoupled S6 gate. For simplicity, in the cartoon voltage sensor activation is depicted by an upward motion of S3S4 with no further conformational changes in S1-S4. (B) Ribbon representation of Kv1.2 crystal structure data (Long et al., 2005a). Two opposite α-subunits are shown from the side (angle of view parallel to the plasma membrane). Cytoplasmic portions preceding S1 have been omitted. The S4-S5 linker, including the initial S5 segment and the distal portion of S6 (orange in A and B), known to be critically involved in voltage-dependent gate opening, were examined here by scanning mutagenesis of Kv4.2.
|
|
Figure 4. Scanning mutagenesis in Kv4.2 S4S5 and S6, and functional expression of mutants. (A) Alignment of Shaker, rat Kv1.2, rat Kv2.1, rat Kv3.1, and human Kv4.2 amino acid sequences in the S4-S5 linker, including the initial part of S5 (S4S5; left) and the distal S6 segment (S6; right). Complementary Shaker sequences in S4S5 and S6 shown by Lu et al. (2002) to be essential for voltage control over the gate are depicted in white letters. Kv4.2 scanning mutagenesis extended from Gly 309 to Phe 329 in S4S5 and from Val 379 to Tyr 413 in S6. Mutations of Gly 314, Leu 327, Phe 329, Ile 398, Pro 403, Ile 405, and Val 406 (red) to alanine produced channels with currents, which could not be distinguished from endogenous oocyte conductances. In previous studies, Cys 320 in Kv4.2 (triangle) or a homologous cysteine in other Kv4 channels has been mutated to serine, residues homologous to Val 402 and Val 404 in Kv4.2 (triangles) to isoleucine (see Materials and methods). The Kv4.2 residues Glu 323 in S4S5 and Val 404 in S6 (filled circles) are homologous to the Shaker residues Glu 395 and Val 476, which have been shown to be critically involved in voltage-dependent gate opening (Yifrach and MacKinnon, 2002). (B) Current families obtained for Kv4.2 wild-type (wt) and individual point mutants with voltage protocols as in Fig. 1 A. Here, only responses to test potentials between −80 and +60 mV in 20-mV increments are shown, according to the illustrated voltage steps (top left). Note the small amplitudes and apparent positive shift in voltage dependence of activation for L310A. (C) Current families obtained for individual S6 point mutants with the same voltage protocols.
|
|
Figure 5. Analysis of voltage dependence of activation and inactivation, and inactivation kinetics for individual point mutants in S4S5 and S6. (A) Analysis for the S4S5 mutants E323A (triangles; n = 8), C320A (circles; n = 5), F326A (squares; n = 5), and L310A (inverted triangles; n = 3 for inactivation and n = 4 for activation). (Left) Steady-state inactivation (filled symbols; variable offset first-order Boltzmann function) and activation curves (open symbols; fourth-order Boltzmann function). The corresponding results for Kv4.2 wild-type (see also Fig. 1 C) are indicated as dotted lines without symbols. Note the large offset in steady-state availability for F326A and the extreme positive shift of the L310A activation curve. (Middle) Onset of low-voltage inactivation analyzed with a single-exponential function (n = 5–6 independent experiments). Wild-type data indicated as a dotted line without symbols. Inactivation kinetics were not analyzed for L310A. (Right) Recovery from inactivation analyzed with a single-exponential function. Wild-type data indicated as a dotted line without symbols. Note that in all shown S4S5 mutants, the recovery from inactivation was accelerated (n = 5–6). (B) Same analysis for the S6 mutants V404A (triangles; n = 4–7), S410A (circles; n = 5), and N408A (squares; n = 5–7). As in A, dotted lines without symbols represent wild-type data. Note that the strong negative shift of the V404A inactivation curve is accompanied by a positive shift of the V404A activation curve.
|
|
Figure 6. Transition rates and apparent affinity for the closed-inactivated state of wild-type and mutant Kv4.2 channels. From the time constants of low-voltage inactivation, τL, and the fractions of non-inactivated channels f at −65 mV, on-rates (kon,-65; A) and off-rates (koff,-65; B) were calculated for Kv4.2 wild-type (wt) and each individual point mutant of S4S5 and S6. Data obtained with the mutants C320S and V402I:V404I (extremely small kon,-65) are included for comparison. Vertical dotted lines in A and B represent wild-type data (wt). (C) The mutation-induced changes in apparent affinity for the closed-inactivated state in Kv4.2 channels relative to wild-type are expressed as ln (Kci/Kci,wt). Note that the effects of the C320A and the C320S mutation differ. Vertical dotted lines indicate ∣ ln (Kci/Kci,wt) ∣ = 1; n = 3–8 independent determinations.
|
|
Figure 7. Functional expression and low-voltage inactivation analysis of Kv4.2 double mutants. (A) Current families obtained for double point mutants, which contain either E323A (left) or S322A (right) in combination with V404A, S407A, N408A, or I412A (only responses to test potentials between −80 and +60 mV in 20-mV increments are shown). (B) Analysis of onset of low-voltage inactivation for the double mutants E323A:V404A (left; n = 5) and E323A:S407A (right; n = 5). Wild-type data are indicated by dotted lines, and data from corresponding individual mutants are shown in light and dark gray, as indicated. Note the apparent additivity of the effects induced by E323A and S407A.
|
|
Figure 8. Double mutant cycle analysis and putative spatial orientation of corresponding S4S5 and S6 amino acid side chains in Kv4.2. (A) Effects of the studied double mutations on the apparent affinity for the closed-inactivated state in Kv4.2 channels expressed as ln(Kci/Kci,wt) and depicted as black bars. Effects of the corresponding individual mutations are indicated as light and dark gray bars. Note that in strong contrast, especially to the combinations containing V404A, additive effects are apparent for the combinations containing S407A. (B) Thermodynamic coupling for the pairs of mutations is quantified as ln Ω. The double mutant cycle analysis indicates strong coupling for the amino acid pairs E323/V404 and S322/V404, whereas no coupling is seen for the amino acid pairs E323/S407 and S322/S407; n = 5–8 independent determinations. (C) Homology modeling of the amino acid residues involved in the double mutant cycle analysis based on Kv1.2 crystal structure data (Long et al., 2005a; zoomed view of the left α-subunit in Fig. 3 B). Orange, scanning regions; green, Kv4.2 S4S5 residues Glu 323 and Ser 322; dark blue, Kv4.2 S6 residues Val 404, Asn 408, and Ile 412. The Kv4.2 S6 residue Ser 407, which in accordance with its low coupling factors points in a different direction, is colored light blue.
|
|
Figure 9. Selective redox sensitivity of a Kv4.2 E323C:V404C double mutant. (A; left) Redox modulation of current amplitude and kinetics for the N-terminally truncated double-cysteine mutant Kv4.2Δ2-10:E323C:V404C coexpressed with KChIP2 in a Xenopus oocyte. Application of 10 mM DTT and 2 mM tbHO2 indicated by horizontal black and gray bars, respectively. (Inset) Current kinetics at time points 1, 2, and 3 during the experiment (normalized and averaged traces from five oocytes; gray shadings represent error bars; horizontal scale bar, 50 ms). (Right) Mean time constants of outward current decay before DTT application (point 1, control; τA = 364 ± 8 ms), 10–15 min after switching to DTT (point 2; τA = 177 ± 13 ms), and when the amplitude had returned to control level (indicated by horizontal dotted lines in A–D), usually after 5–10 min in tbHO2 (point 3; τA = 238 ± 25 ms; n = 5). (B) Redox experiment with an oocyte expressing Kv4.2Δ2-10 in the presence of KChIP2 (control, τA = 180 ± 10 ms; DTT, τA = 139 ± 4 ms; tbHO2, τA = 146 ± 2 ms; n = 3). (C) Redox experiment with an oocyte expressing the single-point mutant Kv4.2Δ2-10:E323C in the presence of KChIP2 (control, τA = 176 ± 11 ms; DTT, τA = 149 ± 19 ms; tbHO2, τA = 135 ± 19 ms; n = 3). (D) Redox experiment with an oocyte expressing the single-point mutant Kv4.2Δ2-10:V404C in the presence of KChIP2 (control, τA = 204 ± 8 ms; DTT, τA = 184 ± 6 ms; tbHO2, τA = 193 ± 4 ms; n = 3). Note that only the E323C:V404C double mutant shows selective redox modulation of current amplitude and kinetics, indicative of disulfide bridge formation between Cys 323 and Cys 404.
|
|
Figure 10. Redox modulation of low voltage–induced inactivation. Kv4.2 mutants with substituted cysteines were coexpressed with both KChIP2 and DPP6. All recordings were started in a DTT-containing bath solution. (A) Peak current amplitude increase (DTT 1), suppression (tbHO2), and recovery (DTT 2) for a Xenopus oocyte expressing Kv4.2Δ2-10:E323C:V404C + KChIP2 + DPP6. (B; left) Amplitude suppression after 10–15 min in tbHO2 and recovery for three oocytes expressing the E323C:V404C double mutant. (Right) Averaged data from three oocytes showing the onset of low-voltage inactivation at −65 mV in DTT 1 (black symbols) and tbHO2 (gray symbols). Fitting curves represent double-exponential functions (DTT 1, τL,fast = 0.11 ± 0.01 s; tbHO2, τL,fast = 0.27 ± 0.04 s; P = 0.0837, paired t test). Open symbols represent individual datasets from two out of three oocytes tested in DTT 2 (double-exponential onset kinetics indicated by dotted fitting curves; τL,fast = 0.13 and 0.16 s, respectively). (C; left) Peak current amplitudes during redox experiments, like the one shown in A, for seven oocytes expressing the E323C single mutant with KChIP2 and DPP6. (Right) Onset kinetics of low-voltage inactivation. Fitting curves represent single-exponential functions (DTT 1, τL = 0.41 ± 0.06 s; tbHO2, τL = 0.47 ± 0.07 s; DTT 2, τL = 0.43 ± 0.06 s; n = 7). (D) Peak current amplitudes and onset kinetics of low-voltage inactivation for eight oocytes expressing the V404C single mutant with KChIP2 and DPP6. (Right) Onset kinetics of low-voltage inactivation. Fitting curves represent single-exponential functions (DTT 1, τL = 0.56 ± 0.05 s; tbHO2, τL = 0.55 ± 0.07 s; DTT 2, τL = 0.50 ± 0.05 s; n = 8).
|
