Functional characterization of the C-terminus of the human ether-à-go-go-related gene K(+) channel (HERG)

Abstract

1. In the present study the functional role of the C-terminus of the human ether-à-go-go-related gene K(+) channel HERG was investigated using a series of C-terminal deletion constructs expressed in Xenopus oocytes. 2. Constructs with deletions of 311 or more amino acid residues failed to form functional channels. Truncation by 215 amino acid residues or fewer had no discernable effects on channel activity. Truncation by 236 or 278 amino acid residues accelerated deactivation, and caused a faster recovery from inactivation. 3. In high extracellular K(+), channel deactivation of HERG results from the binding of the N-terminus to a site within the pore. This slows channel deactivation by a knock-off mechanism. Here, it was shown that C-terminal deletions also abolished this effect of high extracellular K(+). Mutants containing deletions in both the N- and C-termini deactivated with rates similar to those observed in individual deletion mutants. 4. In contrast, experiments with double-deletion constructs showed additive effects of the N- and C-termini on the voltage dependence of activation, and on the kinetics of inactivation and recovery from inactivation. The reduction of inactivation in these mutants contributed to an increase in peak current amplitude. 5. These results indicate that residues within the C-terminus of HERG play a role in channel expression as well as in most aspects of channel gating. The regulation of channel deactivation is likely to be mediated by an interaction with the N-terminus, but the regulation of the voltage dependence of activation, and of rate processes associated with inactivation, does not require the N-terminus.

Figure 1. HERG C-terminal and HERG N- and C-terminal deletions

A, schematic representation of HERG C-terminal deletions. B, schematic representation of K⁺ channel (HERG) structure. C, schematic representation of HERG N- and C-terminal double-deletion constructs.

Figure 5. Effect of high K⁺ concentration on deactivation kinetics of WT, HERGCΔ236 and HERGCΔ278 channels

Deactivation kinetics were studied as in Fig. 3. A, high K⁺ accelerated the deactivation of WT channels. In this set of traces cells were held at −80 mV depolarized to +40 mV for 1 s and subsequently hyperpolarized to −100 mV for 3 s. B, fast and slow deactivation plots for WT HERG + 5 mm K⁺ (▴) and WT HERG + 300 mm K⁺ (▵), indicating that high K⁺ accelerated deactivation of WT HERG (n = 13, P = 0.011 for fast deactivation and P = 0.17 for slow deactivation). C, fast and slow deactivation plots for HERGCΔ236 + 300 mm K⁺ (□) and HERGCΔ236 + 5 mm K⁺ (□), indicating that high K⁺ did not alter the deactivation (n = 1, P = 0.41 for fast deactivation and P = 0.88 for slow deactivation). D, fast and slow deactivation plots for HERGCΔ278 + 300 mm K⁺ (^) and HERGCΔ278 + 5 mm K⁺ (•), indicating that high K⁺ did not alter deactivation (n = 11, P = 0.42 for fast deactivation and P = 0.33 for slow deactivation).

Figure 2. I-V and G-V relationships of HERG C-terminal deletion mutants

A, current traces recorded from HERG C-terminal deletion mutants. In these experiments cells were held at −80 mV and voltage commands were given from −80 to +70 mV for 1 s in steps of 10 mV and each step was followed by a step to −100 mV. B, peak amplitude I-V plots for HERGCΔ163 to HERGCΔ278 (symbols are indicated in figure). Deletion of amino acid residues from the C-terminus reduced the amplitude of HERG currents. One-way ANOVA showed that the mean currents of these constructs at each voltage were significantly different at the 0.05 level (n = 24, P = 0.0023). Deletion of more than 300 amino acid residues from the C-terminus resulted in no detectable current. C, normalized I-V plot for HERGCΔ163 to HERGCΔ278. D, normalized G-V plot for HERG C-terminal deletion mutants. Deletion of amino acid residues from the C-terminus of HERG had little effect on the slope of the steady-state activation curve. Slope (mV, mean ±s.e.m.) and V_1/2 (mV, mean ±s.e.m.) values were 10.1 ± 0.7, −33.9 ± 0.8 for WT; 14.3 ± 2.1, −37.9 ± 2.6 for HERGCΔ163; 11.1 ± 1.0, −36.5 ± 1.2 for HERGCΔ215; 13.3 ± 1.8, −36.1 ± 2.1 for HERGCΔ236; and 9.5 ± 0.46, −27.7 ± 0.5 for HERGCΔ278. One-way ANOVA showed that the voltage dependences were not significantly different at the 0.05 level (n = 16 for each construct, P = 0.999).

Figure 7. I-V and G-V relationships of HERG double-deletion mutants

A, family of current traces recorded from HERG N-terminal deletion and N- and C-terminal double-deletion mutants with different test potentials. Outward currents were recorded in response to a voltage protocol in which commands were given from −80 to +70 mV for 1 s in steps of 10 mV from a holding potential of −80 mV; repolarization to −100 mV elicited inward tail currents. B, I-V relationships indicate that two double-deletion mutants, HERGNΔ354CΔ236 and HERGNΔ354CΔ278, had larger outward currents than WT channels (n = 16, P = 0.000003 for HERGNΔ354CΔ236 and HERGNΔ354CΔ278, and P = 0.000001 for the others). C, G-V relationships indicate that deletion of residues from the N-terminus and both the N- and C-termini of HERG shifted the activation to more positive voltages (n = 13, F= 0.747, P = 0.6). Slope (mV, mean ±s.e.m.) and V_1/2 (mV, mean ±s.e.m.) values were 10.1 ± 0.7, −33.9 ± 0.8 for WT HERG; 16.1 ± 2.5, −22.1 ± 2.4 for HERGNΔ12CΔ236; 17.0 ± 1.7, +2.6 ± 1.7 for HERGNΔ354CΔ236; 16.5 ± 1.9, −11.6 ± 1.9 for HERGNΔ12CΔ278; 36.4 ± 10.7, −2.6 ± 7 for HERGNΔ354CΔ278; 26.7 ± 4.8, −5.9 ± 3.9 for HERGNΔ12; and 13.3 ± 2.7, −26 ± 3 for HERGNΔ354.

Figure 4. Activation and inactivation of HERG C-terminal deletion mutants

A, activation kinetics. Normalized tail currents were plotted versus activation pulse duration for WT HERG, HERGCΔ236 and HERGCΔ278 at +40 mV (only two fast deactivation phenotypes were measured). This protocol consisted of depolarizing steps, each followed by repolarization to −100 mV, given in successive increments of 10 ms at +40 mV. Slope (ms, mean ±s.e.m.) and time constant (ms, mean ±s.e.m.) values were −1.17 ± 0.05, 87.1 ± 5.9 for WT; −0.97 ± 0.02, 76.9 ± 4.9 for HERGCΔ236; and −1.06 ± 0.03, 69.0 ± 3.0 for HERGCΔ278. Activation time constants were not significantly different at the 0.05 level (n = 14, P = 0.764). B, in these experiments, channels were fully activated (and inactivated) by a pulse to +60 mV for 5 s. During a subsequent pulse to −80 mV channels recovered from inactivation. Then, prior to deactivation, a range of depolarizing potentials from +60 to −20 mV were given and channels re-entered the inactivated state. Inactivation time constants of HERG C-terminal deletion mutants were plotted versus voltage and were not different from those of WT channels at the 0.05 level (n = 12, P = 0.901).

Figure 8. Activation kinetics of double-deletion mutants

Activation rates of double-deletion mutants were significantly slower than those of WT channels and those of N- and C-terminal single-deletion mutants. A, normalized tail currents were plotted versus activation pulse duration for WT HERG, HERGNΔ12CΔ236, HERGNΔ12CΔ278, HERGNΔ354CΔ236 and HERGNΔ354CΔ278 at +40 mV (symbols indicated in figure). Activation time constants were fitted with a single-exponential function (see Fig. 4 for details; n = 17, P = 0.408). B, activation time constants at +40 mV for WT HERG, HERGNΔ12 and HERGNΔ354 are shown (n = 17, P = 0.044). Time constants were calculated from exponential fits (symbols indicated in figure). C, activation time constants of WT HERG, HERGNΔ12CΔ236, HERGNΔ12CΔ278, HERGNΔ354CΔ236, HERGNΔ354CΔ236, HERGNΔ12 and HERGNΔ354 are summarized in this bar graph (n = 17). D, slope values of activation curves at +40 mV for the constructs above (n = 17).

Figure 3. Deactivation kinetics of HERG C-terminal deletion mutants

A, deactivation traces recorded from HERG C-terminal deletion mutants, indicating that the two shortest mutants, HERGCΔ236 and HERGCΔ278, had significantly faster deactivation rates than those of WT channels. In this experiment cells were depolarized by activating channels from rest (-80 mV) with a 1 s pulse to +20 mV and measuring the closure of the channels during a range of pulses to 100 mV for 3 s. B, plot of fast deactivation time constants versus voltage for C-terminal deletion mutants (at −120 mV they were 30.4 ± 9 ms for WT, 18.8 ± 9 ms for HERGCΔ163, 17.8 ± 7 ms for HERGCΔ215, 6.01 ± 0.8 ms for HERGCΔ236 and 7 ± 2 ms HERGCΔ278. One-way ANOVA showed that the mean currents of HERGCΔ236 and HERGCΔ278 at each voltage were significantly different from WT at the 0.05 level (n = 13, P = 0.0008) whereas they were not significantly different for HERGCΔ163 and HERGCΔ215 (n = 13, P = 0.166). C, plot of slow deactivation time constants versus voltage for C-terminal deletion mutants. One-way ANOVA showed that the mean currents of HERGCΔ236 and HERGCΔ278 at each voltage were significantly different from WT at the 0.05 level (n = 13, P = 0.045) whereas they were not significantly different for HERGCΔ163 and HERGCΔ215 (n = 13, P = 0.93). Both time constants were significantly shorter at the 0.05 level in the deletion mutants (at −120 mV they were 192 ± 40 ms for WT, 188 ± 54 ms for HERGCΔ163, 185 ± 42.5 ms for HERGCΔ215, 58 ± 16 ms for HERGCΔ236 and 56 ± 9 ms for HERGCΔ278).

Figure 6. Deactivation kinetics of double-deletion mutants

Deactivation in double-deletion mutants was not faster than in individual N-terminal and C-terminal deletion mutants. A, deactivation current traces from HERG C-terminal, N-terminal and double-deletion mutants. In this set of traces cells were held at −80 mV, depolarized to +20 mV for 1 s and subsequently hyperpolarized to −120 mV for 3 s. B, plot of fast deactivation time constants versus voltage for C-terminal, N-terminal and double-deletion mutants. Symbols for plots are indicated in the figure (n = 16, P = 0.383). C, plot of slow deactivation time constants versus voltage for C-terminal, N-terminal and double-deletion mutants. Symbols as in B (n= 16, P = 0.773).

Figure 9. Inactivation kinetics of double-deletion mutants

A, family of current traces recorded from WT HERG, HERGNΔ12CΔ236, HERGNΔ12CΔ278, HERGNΔ354CΔ236, HERGNΔ354CΔ278, HERGNΔ12 and HERGNΔ354 are shown. Inactivation pulse protocol consisted of a pulse to +60 mV for 5 s with a subsequent pulse to −80 mV and a range of depolarizing potentials from +60 to −20 mV. B, inactivation time constants of WT HERG, HERGNΔ12CΔ236, HERGNΔ12CΔ278, HERGNΔ354CΔ236, HERGNΔ354CΔ278, HERGNΔ12 and HERGNΔ354 were plotted against test voltage (n = 20, P = 0.0004). Symbols are defined in the figure. Inactivation was slower in deletion mutants than in WT channels.

Figure 10. Recovery from inactivation in all deletion mutants

A, recovery from inactivation current traces. Pulse protocol consisted of a 1 s pulse to +60 mV followed by a pulse to −30 mV. B, time constants of recovery plotted versus test voltage for WT HERG, C-terminal deletion mutants HERGCΔ236 and HERGCΔ278, and N-terminal deletion mutants HERGNΔ12 and HERGNΔ354. Symbols are indicated in the figure (n = 24, P = 0). C, recovery plots for HERGNΔ354CΔ236, HERGNΔ354CΔ278, HERGCΔ236, HERGCΔ278 and HERGNΔ354 (n = 24, P = 0). All deletion mutants had significantly faster recovery than WT channels.