Modification of hERG1 channel gating by Cd2+

Abstract

Each of the four subunits in a voltage-gated potassium channel has a voltage sensor domain (VSD) that is formed by four transmembrane helical segments (S1-S4). In response to changes in membrane potential, intramembrane displacement of basic residues in S4 produces a gating current. As S4 moves through the membrane, its basic residues also form sequential electrostatic interactions with acidic residues in immobile regions of the S2 and S3 segments. Transition metal cations interact with these same acidic residues and modify channel gating. In human ether-á-go-go-related gene type 1 (hERG1) channels, Cd(2+) coordinated by D456 and D460 in S2 and D509 in S3 induces a positive shift in the voltage dependence of activation of ionic currents. Here, we characterize the effects of Cd(2+) on hERG1 gating currents in Xenopus oocytes using the cut-open Vaseline gap technique. Cd(2+) shifted the half-point (V(1/2)) for the voltage dependence of the OFF gating charge-voltage (Q(OFF)-V) relationship with an EC(50) of 171 microM; at 0.3 mM, V(1/2) was shifted by +50 mV. Cd(2+) also induced an as of yet unrecognized small outward current (I(Cd-out)) upon repolarization in a concentration- and voltage-dependent manner. We propose that Cd(2+) and Arg residues in the S4 segment compete for interaction with acidic residues in S2 and S3 segments, and that the initial inward movement of S4 associated with membrane repolarization displaces Cd(2+) in an outward direction to produce I(Cd-out). Co(2+), Zn(2+), and La(3+) at concentrations that caused approximately +35-mV shifts in the Q(OFF)-V relationship did not induce a current similar to I(Cd-out), suggesting that the binding site for these cations or their competition with basic residues in S4 differs from Cd(2+). New Markov models of hERG1 channels were developed that describe gating currents as a noncooperative two-phase process of the VSD and can account for changes in these currents caused by extracellular Cd(2+).

Figure 1.

Effect of Cd²⁺ on hERG1 channel gating currents. (A) Currents recorded at a V_t of +60 mV using the COVG technique from an uninjected oocyte before (control) and 10 min after bath application of 30 µM Cd²⁺. Recordings were made with the same solutions used to record gating currents in oocytes expressing hERG1 channels. (B and C) Gating currents recorded at the indicated V_t before and after treatment of the same oocyte with 30 µM Cd²⁺. Bottom panels illustrate voltage clamp protocol used to elicit gating currents shown in panels A–C.

Figure 2.

Extracellular Cd²⁺ shifts the Q-V relationship for hERG1 to more positive potentials. (A) Normalized Q_OFF is plotted as a function of V_t and [Cd²⁺]_e. Averaged data for control and each [Cd²⁺]_e were fitted to a Boltzmann function (smooth curves). (B) Plot of change in V_1/2 for Q-V and G-V relationships as a function of [Cd²⁺]_e. Data were fitted to a Hill equation to determine EC₅₀ and Hill coefficient (n_H). ΔV_1/2max was set at +90 mV (Fernandez et al., 2005). For Q-V relationship, EC₅₀ = 171 ± 13 µM; n_H = 0.53 ± 0.03. For G-V relationship, EC₅₀ = 416 ± 20 µM; n_H = 0.57 ± 0.02. The number of oocytes used to calculate the average data is indicated next to each data point. (C) [Cd²⁺]_e-dependent effects on V_1/2 for Q-V relationship. (D) [Cd²⁺]_e-dependent effects on effective valence, z, of Boltzmann function for Q-V relationship. For A, C, and D, the number of oocytes was as follows: control, 20; 3 µM Cd²⁺ , 8; 10 µM Cd²⁺, 7; 30 µM Cd²⁺, 12; 100 µM Cd²⁺, 10; 300 µM Cd²⁺, 14.

Figure 3.

Cd²⁺ reduces the fast and slow components of Ig_ON and induces the appearance of I_Cd-out, a brief initial outward component that precedes Ig_OFF. (A) Superimposed traces for gating currents measured before and after treatment of an oocyte with 30 µM Cd²⁺. Top panel shows pulse protocol. (B) Ig_ON shown on an expanded time scale. (C) Ig_OFF shown on an expanded time scale.

Figure 4.

Gating currents measured without using p/−8 leak and capacitance current subtraction. (A) Unsubtracted currents recorded under control ionic conditions using pulse protocol illustrated in top panel. (B) Unsubtracted currents for the same oocyte in the presence of 30 µM Cd²⁺. (C) Q_OFF-V relationships for gating currents recorded in the absence and presence of 30 µM Cd²⁺. Data were fitted to a Boltzmann function (smooth curves). In control conditions, V_1/2 = −22.1 mV and z = 2.29. In the presence of Cd²⁺, V_1/2 = +7.8 mV and z = 1.52. (D) Superimposed traces of unsubtracted Ig_OFF for control (black traces) and 30 µM Cd²⁺ (red traces) at −110 mV after activating pulses were applied to the indicated V_t. (E) Superimposed traces for p/−8 subtracted (blue) and unsubtracted currents (red) in the presence of 30 µM Cd²⁺ for a V_t of +60 mV. Note that I_Cd-out is approximately the same, indicating that it is not an artifact of the on-line leak subtraction protocol.

Figure 5.

Effect of [Cd²⁺]_e on the magnitude of I_Cd-out. (A) Ig_OFF in the presence of 30 µM Cd²⁺. Arrows indicate zero current level. Gray bar indicates region of Ig_OFF integrated to obtain total Q_OFF. (B) I_Cd-out shown in A, plotted on an expanded time scale. Gray bar indicates the region of Ig_OFF integrated to obtain Q_Cd-out. (C) Integral of Ig_OFF. Inset shows Q_OFF at greater resolution. (D) Plot of Q_Cd-out/Q_OFF total as a function of V_t and [Cd²⁺]_e. Also shown is a plot of Q_Cd-out versus V_t for control conditions. (E) Plot of Q_Cd-out/Q_OFF total as a function of V_t for 100 µM Cd²⁺ and fitted to a Boltzmann function (smooth curve): V_1/2 = 27 mV; slope factor = 11.0 mV; maximum value of Q_Cd-out/Q_OFF total = 0.064.

Figure 6.

Kinetics of Q_OFF. (A) Gating currents elicited under control conditions. Top panel is a plot of gating currents in response to voltage pulse protocol shown in the middle panel. Bottom panel shows the first integral of Ig_OFF traces depicted in top panel. (B) Gating currents, pulse protocol, and integration of currents are shown for the same oocyte after bath application of 30 mM Cd²⁺. (C) Plot of total Q_OFF versus pulse duration determined before (filled squares) and after treatment of an oocyte with Cd²⁺ (open circles). Smooth curves are best fits for single-exponential function (τ = 29 ms for control; τ = 100 ms for Cd²⁺). (D) Plot of Q_Cd-out in the presence of 30 µM Cd²⁺ versus pulse duration. Data for pulse durations >70 ms (filled circles) were fitted with a single-exponential function (τ = 95 ms). Thus, after accounting for a delay caused by overlap of the faster rising phase of Ig_OFF for short pulse durations, the time course of the development of Q_Cd-out matches that of Q_OFF.

Figure 7.

Elevated concentration of Ca²⁺ causes a positive shift in the voltage dependence of the Q_OFF-V relationship for hERG1, but it does not induce a transient outward component in Ig_OFF. (A) Q_OFF-V relationships for hERG1 measured under control conditions ([Ca²⁺]_e = 2 mM) and after the elevation of [Ca²⁺]_e to 10 mM (n = 5). Averaged data were fitted to a Boltzmann function (smooth curves). 2 mM Ca²⁺: V_1/2 = −19.5 ± 0.5 mV and z = 1.87 ± 0.05; 10 mM Ca²⁺: V_1/2 = 6.8 ± 1.9 mV and z = 1.42 ± 0.06. (B) Ig traces recorded in the presence of 2 mM Ca²⁺ (control) and 10 mM Ca²⁺ at a test potential of +60 mV. Ig_OFF was measured at −110 mV. (C) Ig_OFF traces from B are shown at expanded time (initial 5 ms) and amplitude scales. (D) Gating currents of WT hERG1 channels measured during 300-ms pulses to +40 mV with a holding and return potential of −110 mV. Q_OFF-V relationships are shown in G. Extracellular concentrations of Ca²⁺ and Cd²⁺ are indicated to the left of the current traces. (E) Expanded view of initial Ig_OFF for the traces shown in D. (F) Currents recorded from an uninjected oocyte using the same pulse protocol as in D with the indicated concentrations of extracellular divalent cations. (G) Q_OFF-V relationship determined in a single oocyte in the presence of the indicated concentrations of extracellular Ca²⁺ and Cd²⁺. For 2 mM Ca²⁺: Q_OFF-max = −8.0 nC, V_1/2 = −19.6 mV, and z = 2.2. For 10 mM Ca²⁺: Q_OFF-max = −12.0 nC, V_1/2 = +7.3 mV, and z = 1.6. For 10 mM Ca²⁺ + 0.1 mM Cd²⁺: Q_OFF-max = −9.3 nC, V_1/2 = +32 mV, and z = 1.1.

Figure 8.

Co²⁺, Zn²⁺, and La³⁺ cause a positive shift in the voltage dependence of Q_OFF-V relationships for hERG1, but they do not induce a transient outward component in Ig_OFF. (A) Q_OFF-V relationships for hERG1 measured under control conditions ([Ca²⁺]_e = 2 mM) and after the addition of 3 mM Co²⁺ (n = 3). Control: V_1/2 = −15.9 ± 0.8 mV and z = 2.20 ± 0.08; Co²⁺: V_1/2 = 19.1 ± 0.9 mV and z = 1.34 ± 0.02. (B) Ig traces recorded in the presence of 2 mM Ca²⁺ (control) and 3 mM Co²⁺ at a test potential of +60 mV. Ig_OFF was measured at −110 mV. (C) Ig_OFF traces from B are shown at expanded time (initial 5 ms) and amplitude scales. (D) Q_OFF-V relationships for hERG1 measured under control conditions ([Ca²⁺]_e = 2 mM) and after the addition of 1 mM Zn²⁺ (n = 3). Control: V_1/2 = −17.5 ± 0.3 mV and z = 1.82 ± 0.03; Zn²⁺: V_1/2 = 16.4 ± 6.3 mV and z = 0.97 ± 0.06. (E) Ig traces recorded in the presence of 2 mM Ca²⁺ (control) and 1 mM Zn²⁺ at a test potential of +60 mV. Ig_OFF was measured at −110 mV. (F) Ig_OFF traces from E are shown at expanded time (initial 5 ms) and amplitude scales. (G) Q_OFF-V relationships for hERG1 measured under control conditions ([Ca²⁺]_e = 2 mM) and after the addition of 10 µM La³⁺ (n = 3). Control: V_1/2 = −20.6 ± 0.4 mV and z = 2.18 ± 0.06; La³⁺: V_1/2 = 19.3 ± 3.9 mV and z = 1.23 ± 0.08. (H) Ig traces recorded in the presence of 2 mM Ca²⁺ (control) and 10 µM La³⁺ at a test potential of +60 mV. Ig_OFF was measured at −110 mV. (I) Ig_OFF traces from H are shown at expanded time (initial 5 ms) and amplitude scales.

Figure 9.

Extracellular Cd²⁺ induces positive shifts in the G-V relationships of D509C and D456C hERG1 channels. (A and B) G-V relationships for D509C hERG1 (A; n = 6) and D456C hERG1 (B; n = 5) channel currents in the absence (control) and presence of 1 mM extracellular Cd²⁺.

Figure 10.

Cd²⁺ enhances I_Cd-out of WT but not D509C or D456C hERG1 channels. (A) Traces for initial 10 ms of Ig_OFF of WT hERG1 channels elicited by return of membrane potential to −110 mV from the indicated V_t. (B and C) Traces for initial 10 ms of Ig_OFF of D509C (B) or D456C hERG1 (C) hERG1 channels elicited by return of membrane potential to −110 mV from the indicated V_t. Current traces for mutant channels are aligned below the traces for WT channels to facilitate comparisons based on measured shifts in the V_1/2 for activation of ionic currents. For all panels, black traces represent control currents, and red traces represent currents after treatment of an oocyte with 0.1 mM (WT channels) or 0.3 mM (mutant channels) CdCl₂.

Figure 11.

RPR260243 has no effect on hERG1 channel gating currents. (A) Gating currents for WT hERG1 channels measured at test potentials of −10 and +40 mV before (black traces) or 25 min after exposure of an oocyte to 30 µM RPR260243 (red traces). For these experiments, the currents were filtered at 5 kHz and digitized at 20 kHz. (B) Average Q_OFF-V relationships determined under control conditions and after the treatment of the same oocytes with 30 µM RPR260243 for 20–25 min (n = 5). Q_OFF was normalized to the peak of the control value for each oocyte. Data were fitted with a Boltzmann function (smooth curves). In control conditions, V_1/2 = −18.6 ± 0.5 mV and z = 1.84 ± 0.01. In the presence of RPR260243, V_1/2 = −19.0 ± 0.5 mV and z = 1.88 ± 0.01.

Figure 12.

Averaged and normalized Ig of WT hERG1 channels measured using p/−8 leak and capacitance current subtraction. (A) Ig in response to test voltages V_t of −60, −30, …+60 mV during 0–300 ms. The membrane voltage is clamped to −110 mV otherwise. (B and C) Ig_ON at 0–10 ms and 10–300 ms, respectively. (D and E) Ig_OFF at 300–310 ms and 310–600 ms, respectively.

Figure 13.

Schematics of Markov model for gating and ionic currents of WT hERG1 channels. The model consists of two components, one for voltage sensing with the states S₀ to S₁₄ (A) and the other for activation and inactivation with the states C₀, C₁, O, and I (B). The coupling between the components is via the rate coefficient α_C0,C1, which is a function of the states S₄, S₈, S₁₁, S₁₃, and S₁₄ in the component for voltage sensing.

Figure 14.

Features of measured and modeled I_g. Q-V_t relationships for Ig_ON (A) and Ig_OFF (B). Time constants–V_t relationships for Ig_ON (C) and Ig_OFF (D). Relationship of exponential fit coefficients to V_t for Ig_ON (E) and Ig_OFF (F).

Figure 15.

Modeled Ig of WT hERG1 channels. (A) Ig in response to the voltage-clamping protocol used in Fig. 12. (B and C) Ig_ON at 0–10 ms and 10–300 ms, respectively. (D and E) Ig_OFF at 300–310 ms and 310–600 ms, respectively. (F) The initial phase (0–1 ms) of Ig_ON is primarily carried by the fast transition between S₀ and S₁. (G) The slow decay of Ig_ON between 2 and 10 ms is caused by a mixture of increasing and decreasing gating currents associated with the slow transitions in the gating model.

Figure 16.

Averaged and normalized Ig of WT hERG1 channels measured in the presence of extracellular Cd²⁺. (A) Ig measured with test voltages V_t of −60, −30,…+60 mV. The membrane voltage is −110 mV otherwise. (B and C) Ig_ON at 0–10 ms and 10–300 ms, respectively. (D and E) Ig_OFF at 300–310 ms and 310–600 ms, respectively.

Figure 17.

Schematics of Markov model for Ig of WT hERG1 channels in the presence of extracellular Cd²⁺. The model extends the description of Ig in Fig. 13 A with states and transitions describing the movement of Cd²⁺. States S₀ to S₁₄ (green) describe hERG1 configurations with Cd²⁺ bound to a site in the VSD. States S₁₅ to S₂₄ (blue), and S₂₅ to S₃₀ (brown), are associated with hERG1 configurations having one and two Cd²⁺ ions moved in the VSD, respectively.

Figure 18.

Features of measured and modeled Ig in the presence of extracellular Cd²⁺. (A) Maximal Ig_ON-V_t relationship. (B) Maximal Ig_OFF-V_t relationship. Time constants–V_t relationships for Ig_ON (C) and Ig_OFF (D). Relationship of exponential fit coefficients to V_t for Ig_ON (E) and Ig_OFF (F).

Figure 19.

Modeled Ig of WT hERG1 channels in the presence of extracellular Cd²⁺. (A) Ig in response to the voltage-clamping protocol used in Fig. 12. (B and C) Ig_ON at 0–10 ms and 10–300 ms, respectively. (D and E) Ig_OFF at 300–310 ms and 310–600 ms, respectively. (F) Gating current (Ig) caused by movement of VSD and current conducted by Cd²⁺ (I_Cd) for a 300-ms voltage step to +60 mV, followed by a return to the holding potential of −110 mV. The sum of these two currents produces the gating currents simulated to occur in the presence of Cd²⁺. (G) Currents underlying the initial 2 ms of Ig_OFF (300–302 ms). Major contributors to initial Ig_OFF are Cd²⁺currents (I_Cd) related to transitions between S₁₉–S₂₅ and S₂₄–S₃₀ (black) and between S₅–S₁₅ and S₁₄–S₂₄ (pink). Currents related to VSD movement (red, green, and blue) have a small contribution to initial Ig_OFF but dominate after decay of I_Cd.

Figure 20.

Schematic showing state-dependent positions of Cd²⁺ and key charged residues in S2–S4 segments of the VSD of a single hERG1 channel subunit. According to this model, Cd²⁺ is displaced in an outward direction by basic residues in the S4 segment as it moves inward in response to membrane repolarization. (Left) VSD at depolarized potentials and Cd²⁺ in its “depolarized” coordination site. (Right) VSD at a negative transmembrane potential when all channels are closed and Cd²⁺ located in a position defined as the “repolarized” coordination site. (Middle) A putative intermediate state of the channel. K1, K525; R2, R528; R3, R531; R4, R534; D1, D456; D2, D460; D3, D509. The yellow circle represents a Cd²⁺ ion.