Voltage-dependent dynamics of the BK channel cytosolic gating ring are coupled to the membrane-embedded voltage sensor

Abstract

In humans, large conductance voltage- and calcium-dependent potassium (BK) channels are regulated allosterically by transmembrane voltage and intracellular Ca²⁺. Divalent cation binding sites reside within the gating ring formed by two Regulator of Conductance of Potassium (RCK) domains per subunit. Using patch-clamp fluorometry, we show that Ca²⁺ binding to the RCK1 domain triggers gating ring rearrangements that depend on transmembrane voltage. Because the gating ring is outside the electric field, this voltage sensitivity must originate from coupling to the voltage-dependent channel opening, the voltage sensor or both. Here we demonstrate that alterations of the voltage sensor, either by mutagenesis or regulation by auxiliary subunits, are paralleled by changes in the voltage dependence of the gating ring movements, whereas modifications of the relative open probability are not. These results strongly suggest that conformational changes of RCK1 domains are specifically coupled to the voltage sensor function during allosteric modulation of BK channels.

Keywords: BK; allosteric modulation; human; ion channels; molecular biophysics; structural biology; xenopus.

Figure 1.. Voltage dependence of gating ring rearrangements is associated to activation of the RCK1 binding site.

G-V (left panels) and E-V curves (right panels) obtained simultaneously at several Ca²⁺concentrations from (a) the BK667CY construct, (b) mutation of the RCK1 high-affinity site (D362A/D367A), (c) mutation of the Ca²⁺ bowl (5D5A), or (d) both (D362A/D367A 5D5A). Note that the voltage dependence of the E signal is only abolished after mutating the RCK1 high-affinity binding site (b) or both (d). Data corresponding to each Ca²⁺ concentration are color-coded as indicated in the legend at the bottom. Solid curves in the G-V graphs represent Boltzmann fits. For reference, grey shadows in (a–d) left panels represent the full range of G-V curves corresponding to non-mutated BK667CY channels from 0 μM Ca²⁺ to 95 µM Ca²⁺ (indicated with colored dashed lines). Data points and error bars represent average ± SEM (n = 3–14, N = 2–8). Part of the data in (a, b and c) are taken from (Miranda et al., 2013) and (Miranda et al., 2016).

Figure 2.. Modification of the voltage dependence of gate opening does not affect the gating ring voltage-dependent conformational changes.

(a) Topology of the BKα subunit where the voltage sensing domain (VSD), Ca²⁺ sensing domain (gating ring, GR) and pore domain (PD) are indicated by colored dashed lines boxes (see main text for a full description). (b) The three BK functional modules (VSD, PD, GR), schematically represented as colored boxes, interact allosterically. (c) Diagram representing the main effect of the F315A mutation, which is the uncoupling of the VSD to the PD. (d) G-V (left panel) and E-V curves (right panel) obtained simultaneously at several Ca²⁺concentrations after mutation of the F315 site to alanine (BK667CY^F315A). It should be noted that the extent of the shifts induced by the mutation are smaller than previously reported (Carrasquel-Ursulaez et al., 2015), which could arise from the different experimental conditions and/or our fluorescent construct. (e) The interaction with the γ1 subunit favors the VSD-PD coupling mechanism (f) G-V (left) and E-V curves (right) of BK667CY α subunits co-expressed with γ1 subunits. In all panels, data corresponding to each Ca²⁺ concentration are color-coded as indicated in the bottom legend. Colored dashed lines represent the G-V and E-V curves corresponding to BK667CYα channels (Miranda et al., 2013; Miranda et al., 2016). The solid curves in the G-V graphs represent Boltzmann fits. The full range of G-V curves from 0 μM Ca²⁺ to 95 µM Ca²⁺ from BK667CY is represented as a grey shadow in left panels (d and f), for reference. Data points and error bars represent average ± SEM (n = 3–8; N = 2–3).

Figure 3.. Co-expression with β subunits.

(a) β1 subunits have been shown to directly regulate VSD function, shifting V_h(j) to more negative values (b) Left panel, G-V curves obtained at several Ca²⁺concentrations after co-expression of BK667CY with the β1 subunit, which induces a leftward shift in the E-V curves obtained simultaneously (right). (c) β3bNβ1 chimeras produce similar effects to β1 on VSD function, since they retain the N-terminal region of β1 (Castillo et al., 2015). (d) G-V (left) and E-V curves (right) of BK667CY α subunits co-expressed with the β3bNβ1 chimera. Data corresponding to each Ca²⁺ concentration are color-coded as indicated in the legend at the bottom. Colored dashed lines represent the G-V and E-V curves corresponding to BK667CYα channels (Miranda et al., 2013; Miranda et al., 2016). The solid curves in the G-V graphs represent Boltzmann fits. The full range of G-V curves from 0 μM Ca²⁺ to 95 µM Ca²⁺ from BK667CY is represented as a grey shadow in left panels (b and d), for reference. Data points and error bars represent average ± SEM (n = 3–10; N = 2–4).

Figure 3—figure supplement 1.. Co-expression with β3b subunits.

(a) Representative currents obtained after applying depolarizing pulses to inside-out patches expressing BK667CYα (left) or BK667CYα + β3b channels, in the presence of 12 μM Ca²⁺. (b) Left panel, G-V curves obtained at several Ca²⁺concentrations after co-expression of BK667CY with β3b subunits, inducing no appreciable changes in the E-V curves obtained simultaneously (right). Note that when β3b is co-expressed with the BK667CY construct, the kinetics of inactivation are different than those observed with wild-type BK channels (Xia et al., 2000). At first glance, it appears as if the off rate of inactivation is largely increased. Note also that there is a substantial current reduction at the tails as blockade increased with very positive potentials, which is not observed with wild-type BK channels. The simplest interpretation is that the insertion of the fluorescent protein interferes with the kinetics of blockade mediated by the β3b NH2-terminal region (Lingle et al., 2001). Understanding this discrepancy will require further study.

Figure 4.. Mutation of charged residues of BK VSD.

VSD activation was altered by mutation of charged residues in the VSD that modify its voltage of half activation, V_h(j) (a) The R210E mutation induces a negative shift of V_h(j)(b) G-V (left panel) and E-V curves (right panel) obtained simultaneously from constructs BK667CY containing the R210E mutation at several Ca²⁺concentrations. (c) The E219R mutation produces a negative shift of V_h(j) (d) G-V (left panel) and E-V curves (right panel) obtained simultaneously from constructs BK667CY containing the E219R mutation at several Ca²⁺concentrations. (e) The R213E mutation induces a large positive shift of V_h(j) values. (f) G-V (left panel) and E-V curves (right panel) obtained simultaneously from constructs BK667CY containing the R213E mutation at several Ca²⁺concentrations. Data corresponding to each Ca²⁺ concentration are color-coded as indicated in the bottom legend. Colored dashed lines represent the G-V and E-V curves corresponding to non-mutated BK667CYα channels (Miranda et al., 2013; Miranda et al., 2016). The solid curves in the G-V graphs represent Boltzmann fits. The full range of G-V curves from 0 μM Ca²⁺ to 95 µM Ca²⁺ from BK667CY is represented as a grey shadow in left panels (b), (d and f), for reference. Data points and error bars represent average ± SEM (n = 4–10; N = 3–4).

Figure 5.. Voltage dependence of gating ring rearrangements after specific activation of RCK1 high-affinity binding site by Cd²⁺.

(a) Effect of the VSD E219R mutation on the selective activation of RCK1 by Cd²⁺. (b) G-V (left panels) and E-V curves (right panels) obtained simultaneously at several Ca²⁺concentrations from constructs BK667CY^E219R. (c) VSD R201Q mutation induces a positive shift of V_h(j) (d) G-V (left panels) and E-V curves (right panels) obtained simultaneously at several Cd²⁺concentrations from constructs BK667CY^R201Q (e) Effect of the F315A mutation on the selective activation of RCK1 by Cd²⁺. (f) G-V (left panels) and E-V curves (right panels) obtained simultaneously at several Cd²⁺ concentrations from constructs BK667CY^F315A. Data corresponding to each Cd²⁺ concentration are color-coded as indicated in the legend at the bottom. Colored dashed lines represent the G-V and E-V curves corresponding to BK667CYα channels (Miranda et al., 2013; Miranda et al., 2016). The solid curves in the G-V graphs represent Boltzmann fits. The full range of G-V curves from 0 μM Cd²⁺ to 100 µM Cd²⁺ corresponding to non-mutated BK667CY is represented as a grey shadow in left panels (b), (d), and (f), for reference. Data points and error bars represent average ± SEM (n = 3–4; N = 2).

Figure 6.. Voltage dependence of gating ring movements triggered by Ba²⁺.

(a) The RCK2 site is selectively activated by Ba²⁺, which additionally induces pore block. (b) FRET efficiency (E) data obtained at several Ba²⁺ concentrations from BK667CY constructs (Miranda et al., 2016). (c) Effect of the VSD R210E mutation after selective activation of the RCK2 binding site by Ba^2+. (d) E-V curves obtained at several Ba²⁺ concentrations from BK667CY^R210E constructs. (e) Effect of the VSD R213E mutation after selective activation of the RCK2 binding site by Ba²⁺. (f) E-V curves obtained at several Ba²⁺ concentrations from BK667CY^R213E constructs. (g) Effect of the F315A mutation after selective activation of the RCK2 binding site by Ba²⁺ (h) E-V curves obtained at several Ba²⁺ concentrations from BK667CY^F315A constructs. Data corresponding to each Ba²⁺ concentration are color-coded according to the legend at the bottom. For reference, the curve corresponding to 100 μM Ba²⁺ from the BK667CY construct shown in (b) is also shown as a colored dashed line in panels (b, d, f and h). Data points and error bars represent average ± SEM (n = 4–6; N = 2–3).

Figure 6—figure supplement 1.. Additional experiments to characterize voltage dependence of gating ring movements triggered by Ba²⁺.

FRET efficiency (E) data obtained at several Ba²⁺ concentrations from: (a) BK667CY constructs including D362A/D367A mutations to knockout the RCK1 binding site (BK667CY^{D362A D367A}). (b) BK667CY^{D362A D367A} with additional VSD R210E mutation (BK667CY^{D362A D367A R210E}). (c) BK667CY^{D362A D367A} with additional VSD E219R mutation (BK667CY^{D362A D367A E219R}). For reference, the curve corresponding to 100 μM Ba²⁺ from the BK667CY construct shown in Figure 6a is also shown as a grey line in panels (a, b, and c). Ba2+ concentrations are indicated in legend of panel a. (d) E-V curves obtained simultaneously at several Ca²⁺concentrations (indicated in the legend) from constructs BK667CY^{D362A D367A} after addition of 100 μM bbTBA. Data points and error bars in all panels represent average ± SEM (n = 3). Data in panel (a) were partly obtained from (Miranda et al., 2016).

Author response image 1.

Author response image 2.

Author response image 3.