Remodelling inactivation gating of Kv4 channels by KChIP1, a small-molecular-weight calcium-binding protein

Abstract

Calcium-binding proteins dubbed KChIPs favour surface expression and modulate inactivation gating of neuronal and cardiac A-type Kv4 channels. To investigate their mechanism of action, Kv4.1 or Kv4.3 were expressed in Xenopus laevis oocytes, either alone or together with KChIP1, and the K+ currents were recorded using the whole-oocyte voltage-clamp and patch-clamp methods. KChIP1 similarly remodels gating of both channels. At positive voltages, KChIP1 slows the early phase of the development of macroscopic inactivation. By contrast, the late phase is accelerated, which allows complete inactivation in < 500 ms. Thus, superimposed traces from control and KChIP1-remodelled currents crossover. KChIP1 also accelerates closed-state inactivation and recovery from inactivation (3- to 5-fold change). The latter effect is dominating and, consequently, the prepulse inactivation curves exhibit depolarizing shifts (DeltaV = 4-12 mV). More favourable closed-state inactivation may also contribute to the overall faster inactivation at positive voltages because Kv4 channels significantly inactivate from the preopen closed state. KChIP1 favours this pathway further by accelerating channel closing. The peak G-V curves are modestly leftward shifted in the presence of KChIP1, but the apparent 'threshold' voltage of current activation remains unaltered. Single Kv4.1 channels exhibited multiple conductance levels that ranged between 1.8 and 5.6 pS in the absence of KChIP1 and between 1.9 and 5.3 pS in its presence. Thus, changes in unitary conductance do not contribute to current upregulation by KChIP1. An allosteric kinetic model explains the kinetic changes by assuming that KChIP1 mainly impairs open-state inactivation, favours channel closing and lowers the energy barrier of closed-state inactivation.

Figure 1. Outward currents mediated by Kv4.1 and Kv4.3 channels

A and B, from a holding potential of −100 mV, control Kv4.1 or Kv4.3 currents were evoked by step depolarizations from −80 to +50 mV, in 10 mV intervals. The interpulse interval was 5 s. Note that Kv4.3 currents exhibit a larger fast phase in the development of inactivation. C and D, currents mediated by Kv4.1 or Kv4.3 channels coexpressed with KChIP1. Note that KChIP1 renders the Kv4.1 and Kv4.3 currents nearly indistinguishable. E and F, Kv4.1 and Kv4.3 currents at 0 and +50 mV (from panels A-D) were scaled and superimposed to compare the development of inactivation. Thin line, control; thick line, coexpressed with KChIP1. The inset shows the first 200 ms of the corresponding traces. Note the crossover of the currents at +50 mV.

Figure 8. Voltage dependence of prepulse inactivation and activation parameters

A and B, relation between the prepulse potential and the normalized peak current (I/I_max). The voltage protocol in these experiments was as described in the legend to Fig. 7, except that the duration of the prepulse was fixed and the prepulse voltage was varied. The prepulse length was 12 s and 15 s for Kv4.1 in the absence and presence of KChIP1, respectively, and 20 s for Kv4.3, both in the absence and presence of KChIP1. Filled and open symbols represent the mean values of the normalized peak current (n = 3–6) in the absence and presence of KChIP1, respectively. The continuous lines are the best-fit first-order Boltzmann functions (plus a constant). The best-fit parameters (V_m = midpoint potential, k = slope factor; and NI = non-inactivating fraction) are shown in the graphs. C and D, peak G-V curves. The corresponding peak conductance-voltage relations (G_p-V_t) derived from currents evoked by the pulse protocol described in Fig. 1 legend. The peak chord conductance (G_p) was calculated from the following relation: G_p= I_p/(V_t - V_r), where I_p is the peak current, V_t is the test potential and V_r is the reversal potential of the current (−93 to −95 mV, as determined from the instantaneous current-voltage relation derived from the corresponding tail current experiments). The instantaneous current-voltage relations are linear over the range of voltages within which the channels activate (not shown). Each symbol represents the mean value of 5–13 experiments. The continuous lines represent the best-fit 4th order Boltzmann functions (Smith-Maxwell et al. 1998). The parameters of these fits are given in the graph (V_m is the midpoint voltage for the activation of 1 subunit and k is the corresponding slope factor). The data are shown normalized to the maximum estimated peak conductance (G_p,Max). E and F, voltage dependence of time-to-peak. Currents were evoked by the pulse protocol described in Fig. 1 legend.

Figure 7. Kinetic analysis of the development of inactivation at −50 mV

To examine the development of closed state inactivation, outward currents were evoked by a 250 ms step depolarization to +50 mV from an increasingly prolonged prepulse to −50 mV (the corresponding control traces were evoked from a holding potential of −100 mV). This voltage activates < 3 % of the total conductance (Fig. 9). The interval between episodes was ≥ 5 s. A and B, time courses of closed-state inactivation at −50 mV. The data points are normalized peak currents evoked as explained above. The normalized values resulted from dividing the peak currents, I, evoked from −50 mV by the peak control current, I_max. Each symbol represents the mean value of 4–5 experiments. The continuous lines are the best-fit exponential curves. For Kv4.1 the derived time constants are 408 ms and 166 ms in the absence and presence of KChIP1, respectively. For Kv4.3 the derived time constants are 2.9 s and 1.1 s in the absence and presence of KChIP1, respectively.

Figure 9. Amplitude analysis of Kv4.1 single channel currents

A, single channel currents evoked by a step depolarization to +90 mV from the indicated holding potentials. The interpulse interval was 1–2 s. The holding potential was adjusted to suppress the total current and allow the recording of unitary currents. The traces shown are 2 sets of 3 and 4 consecutive sweeps. The continuous straight line across the traces indicates the zero current level (closed level). Recordings were low-pass filtered at 2 kHz (8-pole Bessel filter) and digitized at 10 kHz. For display and analysis the records were digitally filtered at 500 Hz. B, ensemble average currents (90 traces from the experiments shown in A). Note that the kinetics of these currents correspond to those of the macroscopic Kv4.1 currents (e.g. Fig. 1). Clearly, in the presence of KChIP1 the channels undergo faster inactivation. C, single-channel current-voltage relations. The mean amplitudes were derived by fitting a sum of Gaussian terms (2–4) to all-point histograms generated from the experiments shown above (n = 7–27 selected traces at the indicated membrane potentials). The continuous lines are linear regressions constrained to cross the x-axis at the estimated reversal potential (Fig. 8 legend; −93 to −95 mV). The derived slope conductances are indicated on the graphs. In the presence of KChIP1 the large conductance (∼5 pS) was not generally observed. Similar results were obtained from voltage ramp protocols.

Figure 10. Kinetic model of Kv4 gating (Scheme 2)

A, Kv4 gating in the absence of KChIP1. This model is based on previously developed models of Kv4 gating (Jerng et al. 1999; Beck & Covarrubias, 2001; Bähring et al. 2001a). B, Kv4 gating in the presence of KChIP1. Thicker arrows represent transitions favoured by KChIP1. A dashed arrow between O and I₅ indicates a less favourable inactivation from the open state. The modelled currents are shown in Fig. 11, assuming the parameter values from Table 1. A relatively small allosteric factor (f = 0.3) allows for a significant coupling between voltage-dependent activation and inactivation from closed states (i.e. inactivation is more likely to occur from late closed states in the activation pathway).

Figure 11. Observed and simulated Kv4.3 currents

A, observed currents replotted from Fig. 1 at intervals of 20 mV (−50, −30, −10, +10, +30 and +50 mV). Normalized and superimposed traces depict control Kv4.3 currents (thin lines) and KChIP1-remodelled Kv4.3 currents (thick lines). B, simulated currents. These simulations assume Scheme 2 and the parameter values from Table 1. The set of parameter values used here was constrained to simultaneously account for the following properties: voltage dependence of activation and inactivation, time-to-peak, recovery from inactivation, the development of inactivation at negative voltages and the overall time course of the tail currents at hyperpolarized voltages. The parameter values (rate constants in s⁻¹) that are changed to account for the effects of KChIP1 on inactivation gating are shown along with the corresponding family of simulated currents.

Figure 2. Kinetic analysis of the development of macroscopic Kv4.1 inactivation

A, voltage dependence of the time constants derived from fitting a sum of 3 exponential terms plus a constant to the development of macroscopic inactivation in the absence of KChIP1; •,τ_FAST, ○, τ_INT and ▴, τ_SLOW (Jerng & Covarrubias, 1997). B, corresponding fractional amplitudes of the exponential terms (e.g. fraction A_FAST = A_FAST / (A_FAST + A_INT + A_SLOW + A_NI)). ▵, the fractional amplitude of the constant term, NI. C, voltage dependence of the time constants derived from fitting a sum of 2 exponential terms plus a constant to the development of macroscopic inactivation in the presence of KChIP1; •, τ_FAST and ○, τ_SLOW. In the latter case, a close examination of the residuals from the fits (comparing 1, 2 or 3 exponential terms; Jerng & Covarrubias, 1997) showed that the most significantly improved fits are obtained when upgrading from 1 to 2 exponential terms. In contrast with the control currents, no appreciably improved fits were obtained when 3 exponential terms were assumed to describe the decay of the currents expressed in the presence of KChIP1. D, corresponding fractional amplitudes of the exponential terms (see B). ▵, the fractional amplitude of the constant term, NI. Note that, in the presence of KChIP1, a new slow time constant dominates the development of inactivation at voltages > +30 mV. Each symbol represents the mean value of 3–6 experiments. Here and in subsequent figures, error bars are not apparent when they are not larger than the symbol.

Figure 3. Kinetic analysis of the development of macroscopic Kv4.3 inactivation

Panels and symbols are as described in Fig. 2. Note that as a result of the remodelling by KChIP1 C and D in Figs 2 (Kv4.1) and 3 (Kv4.3) are very similar. Each symbol represents the mean value of 3–10 experiments.

Figure 4. Kinetic analysis of recovery from inactivation

A-D, Kv4.1 and Kv4.3 currents evoked by a double pulse protocol to investigate the kinetics of recovery from inactivation. The first pulse (from −100 mV to +50 mV) allows activation and complete inactivation of the evoked current and the second shorter pulse (+50 mV) tests the amount of current recovered after an increasingly prolonged interpulse interval (at −100 mV). The interval between episodes was 5 s. E and F, the time course of recovery from inactivation. The data points are normalized peak currents evoked as explained above. The normalized value is the result of dividing the test peak currents, I, by the corresponding control peak current, I_max. Each symbol represents the mean value of 3–5 experiments. Continuous line, the best-fit exponential. The derived time constants are shown in the graph.

Figure 5. Kinetic analysis of Kv4.1 tail current deactivation

A and B, tail currents in the absence and presence of KChIP1, respectively. From a holding potential of −100 mV, a 5 ms pulse to +50 mV was delivered first to activate a significant outward current and a subsequent membrane repolarization evoked the tail current to examine the deactivation kinetics (the tested membrane potentials are indicated in C). The interval between episodes was 5 s. The outward current evoked by the pulse to +50 mV has been clipped to emphasize visualization of the inward tail currents. C, voltage dependence of the fast and slow time constants extracted from double exponential fits to the tail current relaxations. Filled and open symbols represent the time constants in the absence and presence of KChIP1, respectively. The continuous lines describe the voltage dependence of the time constants assuming a simple exponential function (τ_FAST) or an exponential term plus a constant (τ_SLOW). From these fits, the estimated equivalent electronic charges are: 0.3 e₀ (τ_FAST) and 0.6 e₀ (τ_SLOW) in the absence of KChIP1 and 0.4 e₀ (τ_FAST) and 1.0 e₀ (τ_SLOW) in the presence of KChIP1. The constant term (voltage-independent) is 3 ms and 1 ms in the absence and presence of KChIP1, respectively. D, voltage dependence of the amplitude ratios (A_FAST /A_SLOW) extracted from double exponential fits to the tail current relaxations. Filled and open symbols represent the amplitude ratios in the absence and presence of KChIP1, respectively. The symbols in all panels represent the mean values of 4–5 experiments. Note that the fast time constant dominates at hyperpolarized voltages and that, especially in this voltage range (−110 to −140 mV), both time constants of deactivation at are reduced in the presence of KChIP1.

Figure 6. Kinetic analysis of Kv4.3 tail current deactivation

Pulse protocol, panels and symbols are as described in Fig. 5. Note that in the absence of KChIP1 a single exponential function was sufficient to describe the tail currents at hyperpolarized voltages (−110 to −140 mV), but 2 exponential terms are needed to describe them at voltages above the reversal potential (∼ −95 mV). The continuous lines describe the voltage dependence of the time constants assuming a simple exponential function (τ_FAST and τ_SLOW in the absence of KChIP1 and τ_FAST in the presence of KChIP1) or an exponential term plus a constant (τ_SLOW in the presence of KChIP1). The equivalent electronic charges estimated from these fits are: 0.4 e₀ (τ_FAST and τ_SLOW in the absence of KChIP1) and 0.3 e₀ (τ_FAST) and 0.8 e₀ (τ_SLOW) in the presence of KChIP1. The constant term (voltage-independent) is 3 ms (τ_SLOW in the presence of KChIP1). The symbols in all panels represent the mean values of 3–4 experiments. Note that in the presence of KChIP1 a new fast time constant dominates the relaxation of the tail current at hyperpolarized voltages.

Scheme 1

Figure 12. Conceptual state diagram of Kv4 inactivation gating

A, inactivation gating near the open state in the absence of KChIP1. Only 2 opposing subunits of the K⁺ channel tetramer are represented. The N-terminal tetramerization T1 domain is represented as a 2 part dark red block at the base of the channel. In the closed state, C, the channel is activated but still closed. The positive charges in the voltage sensor (blue cylinders) have moved outward (the membrane is depolarized). This closed state precedes the opening of the pore and is inactivation permissive. Note the opening of an internal gate when the channel opens (O). Ions (black dots) enter the channel through lateral internal windows (Miller, 2000) and cross the membrane through the K⁺-selective pore (i.e. mediating an outward current under physiological conditions). The following 2 inactivated states are represented: I_o (from the open state) and I_c (from the preopen closed state). Note that the proximal region of the cytoplasmic C-terminal domains of the tetramer (green ribbons) holds the distal N-termini in place (dark red ribbons tethered to the T1 domain). The N-terminus readily occludes the inner mouth of the conduction pathway when the channel opens. Additionally, the channel inactivates from the preopen inactivation permissive closed state (C). Note a conformational change at the internal mouth of the pore, which partly collapses its internal vestibule. B, inactivation gating near the open state in the presence of KChIP1. Calcium-bound KChIPs (hexagons with red dots) interact with the distal N-terminal moieties. This interaction immobilizes the channel's N-termini, hindering inactivation from the open state (dotted arrow). Consequently, by immobilizing the channel's N-termini away from the internal mouth of the pore KChIP1 favours channel closing and the conformational changes that mediate closed-state inactivation. This inactivation pathway contributes significantly to the development of inactivation at positive voltages because the opening step of Kv4 channels is not strongly forward-biased (Table 1). Effectively, KChIP1 lowers the energy barrier of closed state inactivation by reducing the steric effects of the channel's N-termini.