Characterization of the PCMBS-dependent modification of KCa3.1 channel gating

Abstract

Intermediate conductance, calcium-activated potassium channels are gated by the binding of intracellular Ca(2+) to calmodulin, a Ca(2+)-binding protein that is constitutively associated with the C terminus of the channel. Although previous studies indicated that the pore-lining residues along the C-terminal portion of S6 contribute to the activation mechanism, little is known about whether the nonluminal face of S6 contributes to this process. Here we demonstrate that the sulfhydral reagent, parachloromercuribenze sulfonate (PCMBS), modifies an endogenous cysteine residue predicted to have a nonluminal orientation (Cys(276)) along the sixth transmembrane segment (S6). Modification of Cys(276) manipulates the steady-state and kinetic behavior of the channel by shifting the gating equilibrium toward the open state, resulting in a left shift in apparent Ca(2+) affinity and a slowing in the deactivation process. Using a six-state gating scheme, our analysis shows that PCMBS slows the transition between the open state back to the third closed state. Interpreting this result in the context of the steady-state and kinetic data suggests that PCMBS functions to shift the gating equilibrium toward the open state by disrupting channel closing. In an attempt to understand whether the nonluminal face of S6 participates in the activation mechanism, we conducted a partial tryptophan scan of this region. Substituting a tryptophan for Leu(281) recapitulated the effect on the steady-state and kinetic behavior observed with PCMBS. Considering the predicted nonluminal orientation of Cys(276) and Leu(281), a simple physical interpretation of these results is that the nonluminal face of S6 forms a critical interaction surface mediating the transition into the closed conformation, suggesting the nonluminal C-terminal portion of S6 is allosterically coupled to the activation gate.

Figure 1.

PCMBS increases KCa3.1 steady-state current. Macroscopic current records from KCa3.1 channels heterologously expressed in HEK cells to determine sensitivity to PCMBS. PCMBS (500 µM) was added to inside-out patches excised in 0.3 (A), 0.6 (B), and 10 µM (C) Ca²⁺_i after the establishment of steady-state current level. PCMBS washout lasted at least 1 min to ensure channel activation was due to covalent binding, rather than an indirect interaction between channel and compound. (D) KCa3.1 current record demonstrating that PCMBS cannot activate the channel in the absence of Ca²⁺_i. PCMBS can only increase current when the channel is in the open state. The patch was initially excised in 10 µM Ca²⁺_i, followed by the addition of a Ca²⁺_i-free solution, and then a Ca²⁺_i-free solution+PCMBS (500 µM) was added to the bath solution. PCMBS+Ca²⁺_i-free solution was washed out with 10 µM Ca²⁺_i to reactivate the channel. After reaching a steady-state current level, PCMBS (500 µM) was reapplied to the patch, resulting in channel activation.

Figure 2.

PCMBS increases steady-state P_o(max). Variance analysis of KCa3.1 channel current was used to estimate P_o(max), N, and i in the absence and presence of PCMBS. (A) Representative Ca²⁺ concentration response experiment recorded from inside-out macropatches expressing KCa3.1 channels. Patches were initially excised in 10 µM Ca²⁺_i, followed by a series of Ca²⁺_i concentrations (0, 0.1, 0.15, 0.2, 0.25, 0.3, 0.6, 1.0, and 10 µM) applied for 10-s intervals using a fast solution exchanger. After establishing steady-state current in 10 µM Ca²⁺_i, PCMBS (500 µM) was added to the bath solution until reaching a second, steady-state, current level. (B) Variance analysis to determine P_o(max), N, and i in the absence of PCMBS. Plot of variance (σ²), which was calculated from the portion of the record excluding the time when PCMBS was added, against mean current 〈I〉 as fit with Eq. 1 (solid line), and P_o(max) was calculated using Eq. 2: P_o(max) = 0.61, n = 1,071, and i = 3.5 pA. (C) Variance analysis to determine P_o(max), N, and i in the presence of PCMBS. Plot of σ², which was calculated from the entire record against 〈I〉 was fit with Eq. 1 (solid line), and P_o(max) was calculated using Eq. 2: P_o(max) = 0.91, n = 1,078, and i = 3.4 pA.

Figure 3.

PCMBS shifts apparent Ca²⁺ affinity. Complete Ca²⁺ concentration response experiments were performed to estimate EC₅₀ and the Hill coefficient (h) in the absence and presence of PCMBS. (A) Representative macroscopic current record from an inside-out patch–expressing KCa3.1 channels. The patch was excised in 10 µM Ca²⁺_i followed by a series of Ca²⁺_i concentrations beginning with 0, 0.1, 0.3, 0.6, 1.0, 3.0, and 10 µM Ca²⁺_i , all applied in 10-s intervals using a rapid solution exchanger. After the current reached a steady-state level, PCMBS (500 µM) was added, and a second Ca²⁺_i concentration response experiment was performed using the above Ca²⁺_i concentrations in the presence of PCMBS. (B) Plot of normalized 〈I〉 current against the corresponding Ca²⁺_i for KCa3.1 (■) and KCa3.1+PCMBS (▾) fit with a variation of the Hill equation (see Materials and methods). This analysis gave estimates for KCa3.1 (solid line, n = 57), EC₅₀ = 508 ± 13 nM and h = 2.0 ± 0.2, and KCa3.1+PCMBS (dashed line, n = 6), EC₅₀ = 235 ± 17 nM and h = 2.5 ± 0.2. All experiments were done in pairs, but an alternate set of Ca²⁺_i concentrations (0.1, 0.15, 0.2, 0.25, 0.3, 0.6, 1.0, and 10 µM) were used when estimating EC₅₀ and h for KCa3.1+PCMBS. Error bars represent SEM. The error bars for KCa3.1 are smaller than the symbols (■).

Figure 4.

PCMBS modulates channel kinetics. Activation and deactivation kinetics were estimated with Ca²⁺ jump experiments in the absence and presence of PCMBS. A rapid solution exchanger was used to quickly alternate between a Ca²⁺_i-free solution and one of four Ca²⁺_i solutions: (A) 0.5, (B) 0.7, (C) 1.0, and (D) 10 µM Ca²⁺_i according to the protocol illustrating one complete Ca²⁺ cycle shown in the inset of this figure. 5–10 complete Ca²⁺ cycles were recorded from a single patch in the absence and presence of PCMBS (500 µM). Activation and deactivation sweeps in the absence of PCMBS (black trace) were collected and averaged according to their respective Ca²⁺ concentration and superimposed against the corresponding PCMBS sweep (red trace). Current traces were normalized to the max current level, facilitating comparison of time courses between KCa3.1 and KCa3.1+PCMBS. Activation (E) and deactivation (F) rates were estimated by fitting activation and deactivation records with an exponential function and reported as a time constant (τ-on and τ-off) in the absence (▴) and presence (▵) of PCMBS, see also Table I.

Figure 5.

Homology model of the pore region for the open state of KCa3.1 using the rKv1.2 structure. (A) Representation of the S5-P-helix-S6 region viewed as a profile illustrating the predicted orientation of Cys²⁶⁷, Cys²⁶⁹, Cys²⁷⁷ (all in gray), Cys²⁷⁶ (yellow), and Leu²⁸¹ (red). Only two of the four subunits are shown for clarity. (B) Representation of the S5-P-helix-S6 region from the extracellular entrance showing the predicted nonluminal orientation of Cys²⁷⁶ (yellow) and Leu²⁸¹ (red). The S5 and P-helix are colored green and the S6 helix is colored blue. All residues are illustrated in VDW format. (C) Sequence alignment used for the homology model of KCa3.1 in the open state (rKv1.2) and the closed state (KcsA model illustrated in Fig. 15). Bold font represents homologous amino acids used as a guide to align sequences.

Figure 6.

PCMBS increases channel activation for KCa3.1 C277A. Ca²⁺ concentration response experiments and variance analysis were undertaken to determine whether PCMBS increases C277A channel activation in the same manner observed for the WT channel. (A, left) Representative macroscopic current record from an inside-out patch expressing KCa3.1 C277A channels. All experiments were performed using the Ca²⁺ concentraiton response protocol described in Fig. 3. (A, right) Plot of normalized 〈I〉 current against the corresponding [Ca²⁺]_i for KCa3.1 C277A (■) and KCa3.1 C277A+PCMBS (▾). All analysis was performed according to the protocol described in Fig. 3 to give averages: KCa3.1 C277A (solid line, n = 15), EC₅₀ = 566 ± 24 nM and h = 1.9 ± 0.1, and KCa3.1 C277A+PCMBS (dashed line, n = 7), EC₅₀ = 220 ± 14 nM and h = 1.7 ± 0.2. Error bars represent SEM. (B) Variance analysis as described in Fig. 2 to determine P_o(max), N, and i in the absence of PCMBS: P_o(max) = 0.51, n = 954, and i = 3.5 pA. (C) Variance analysis as described in Fig. 2 to determine P_o(max), N, and i in the presence of PCMBS P_o(max) = 0.70, n = 849, and i = 3.5 pA. Analysis of multiple patches (n = 6) indicates that PCMBS increases P_o(max) from 0.55 ± 0.02 to 0.84 ± 0.03 without increasing N or i (3.8 ± 0.1 pA versus 3.9 ± 0.1 in PCMBS).

Figure 7.

C276A prevents the PCMBS-mediated increase in channel activation. Ca²⁺ concentration response experiments and variance analysis were undertaken to determine whether PCMBS increases C276A channel activation in the same manner observed for the WT channel. (A) Representative macroscopic current record from an inside-out patch expressing KCa3.1 C276A channels. All experiments were performed using the Ca²⁺ concentraiton response protocol described in Fig. 3. (B) Plot of normalized 〈I〉 current against the corresponding Ca²⁺_i for KCa3.1 C276A (■) and KCa3.1 C276A+PCMBS (▾). All analyses were performed according to the protocol described in Fig. 3 to give averages: KCa3.1 C276A (solid line, n = 18), EC₅₀ = 460 ± 27 nM and h = 1.7 ± 0.04, and KCa3.1 C276A+PCMBS (dashed line, n = 9), EC₅₀ = 485 ± 68 nM and h = 1.6 ± 0.04. Error bars represent SEM. (C) Variance analysis as described in Fig. 2 to determine P_o(max), N, and i in the absence of PCMBS: P_o(max) = 0.67, n = 488, and i = 3.3 pA. PCMBS did not increase steady-state current in KCa3.1 C276A; therefore, to obtain an estimate of P_o(PCMBS), P_o had to be extrapolated using the equation where P_o = P_o(max)(I/I_(max)) and P_o(PCMBS) = 0.63. Analysis of multiple patches (n = 6) indicates that PCMBS did not increase P_o(max) (0.68 ± 0.05 in the absence of PCMBS and 0.59 ± 0.05 in PCMBS).

Figure 8.

PCMBS modulates C277A channel kinetics in a manner similar to the WT channel. Activation and deactivation kinetics were estimated with Ca²⁺ jump experiments using Ca²⁺_i concentrations (A) 0.5, (B) 0.7, (C) 1.0, and (D) 10 µM in the absence (black trace) and presence (red trace) of PCMBS as described in Fig. 4. Activation (E) and deactivation (F) rates were estimated by fitting activation and deactivation records with an exponential function and reported as a time constant (τ-on and τ-off) in the absence (▴) and presence (▵) of PCMBS, see also Table I.

Figure 9.

C276A prevents the PCMBS-mediated modulation on channel kinetics. Activation and deactivation kinetics were estimated with Ca²⁺ jump experiments using Ca²⁺_i concentrations 0.5, (A) 0.7, (B) 1.0, (C) and 10 µM (D) in the absence (black trace) and presence (red trace) of PCMBS as described in Fig. 4. Activation (E) and deactivation (F) rates were estimated by fitting activation and deactivation records with an exponential function and reported as a time constant (τ-on and τ-off) in the absence (•) and presence (○) of PCMBS, see also Table I.

Figure 10.

A four-state model can be used to fit the activation and deactivation kinetics for KCa3.1. (A) Gating scheme used to describe the activation and deactivation kinetics of KCa3.1. The six-state model is comprised of four closed states and two open states, with forward transitions between closed states being Ca²⁺ dependent and all other transitions being Ca²⁺ independent. The box outline represents the states that are not necessary to fit the activation and deactivation kinetics as determined through sensitivity analysis (Table S1). (B) Representative activation and deactivation records fit with two variations of the model shown in A. Activation and deactivation currents were recorded and fit according to the protocol described in the Materials and methods sections. The colors refer to the different Ca²⁺ concentrations, with red representing 0.5, blue 0.7, green 1.0, and black 10 µM Ca²⁺. The solid line represents the fit assuming Ca²⁺-dependent rate constants have a linear dependence on Ca²⁺ concentration, k = A · [Ca]. The dashed line represents the fit assuming Ca²⁺-dependent rate constants have a nonlinear dependence on Ca²⁺ concentration, k = A · [Ca]/(B + [Ca]). Rate constants derived from the fit are summarized in Table II. (C) The model can be used to predict the shift in apparent Ca²⁺ affinity with PCMBS. Plot of normalized 〈I〉 current against the corresponding Ca²⁺_i for KCa3.1 and KCa3.1+PCMBS. The symbols represent the experimental data (KCa3.1, ◆, and KCa3.1+PCMBS, ■), which is plotted against the predicted apparent Ca²⁺ affinity (KCa3.1, solid line) and (KCa3.1+PCMBS, dashed line). Model prediction: KCa3.1, EC₅₀ = 540 nM, and KCa3.1+PCMBS, EC₅₀ = 280 nM. Experimental observation: KCa3.1, EC₅₀ = 508 nM, and KCa3.1+PCMBS, EC₅₀ = 235 nM.

Figure 11.

PCMBS slows transition k₅₃, as determined through the constrained version of the model. (A) Representative activation and deactivation records fit with two variations of the model. The dashed line represents the variation that allowed parameters k₁₂, k₂₁, k₂₃, k₃₂, k₃₅, and k₅₃ to be free, and the solid line represents the variation that held parameters k₁₂, k₂₁, k₂₃, k₃₂, and k₃₅ to k_(-PCMBS) values; k₅₃ is the only free parameter, see Results for details. As noted in the previous figure, the Ca²⁺-dependent rate constants were assumed to have a nonlinear dependence on Ca²⁺ concentration, k = A · [Ca]/(B + [Ca]). The colors refer to the different Ca²⁺ concentrations, with red representing 0.5, blue 0.7, green 1.0, and black 10 μM Ca²⁺. (B) Gating scheme detailing that parameter k₅₃ is modulated by PCMBS, as determined through the constrained version of the model.

Figure 12.

Trp²⁸¹ and Trp²⁸² recapitulate the PCMBS-mediated shift in apparent Ca²⁺ affinity. Complete Ca²⁺ concentration response experiments were performed to estimate EC₅₀ and Hill coefficients (h) for those constructs expressing macroscopic currents, as described in Fig. 3. (A) Plot of normalized 〈I〉 current against the corresponding Ca²⁺_i for KCa3.1 C276W (•) and C277W (□). KCa3.1 (▴) and KCa3.1+PCMBS (▵) are included for comparison. KCa3.1 C276W (n = 7), EC₅₀ = 558 ± 50 nM and h = 1.8 ± 0.1; KCa3.1 C277W (n = 6), EC₅₀ = 748 ± 54 nM and h = 1.5 ± 0.1; KCa3.1 (n = 57), EC₅₀ = 508 ± 13 nM and h = 2.0 ± 0.2; and KCa3.1+PCMBS (n = 6), EC₅₀ = 235 ± 17 nM and h = 2.5 ± 0.2. (B) Plot of normalized 〈I〉 current against the corresponding [Ca²⁺]_i for KCa3.1 L281W (■), KCa3.1 L281W+PCMBS (□), and KCa3.1 V282W (○). KCa3.1 (▴) and KCa3.1+PCMBS (▵) are included for comparison. KCa3.1 L281W (n = 12), EC₅₀ = 54 ± 6.4 nM and h = 3.3 ± 0.2; KCa3.1 L281W+PCMBS (n = 3), EC₅₀ = 43 ± 1.2 nM and h = 3.6 ± 0.5; KCa3.1 V282W (n = 12), EC₅₀ = 296 ± 14 nM and h = 1.6 ± 0.1; KCa3.1 (n = 57), EC₅₀ = 508 ± 13 nM and h = 2.0 ± 0.2; and KCa3.1+PCMBS (n = 6), EC₅₀ = 235 ± 17 nM and h = 2.5 ± 0.2. Error bars represent standard SEM. Error bars smaller than the symbols are not visible in the graph.

Figure 13.

Trp²⁸¹ recapitulates the PCMBS-mediated modulation in deactivation kinetics. Activation and deactivation kinetics were estimated with Ca²⁺ jump experiments as described in Fig. 4 using Ca²⁺_i concentrations (A) 0.5, (B) 0.7, (C) 1.0, and (D) 10 µM for L281W (blue trace) and KCa3.1 in the absence (black trace) and presence (red trace) of PCMBS, included for comparison. Activation (E) and deactivation (F) rates were estimated by fitting activation and deactivation records with an exponential function and reported as a time constant (τ-on and τ-off) for KCa3.1 (▴), KCa3.1 PCMBS (▵), L281W (•), L281W PCMBS (○), and V282W (□), see also Table I.

Figure 14.

A four-state model can be used to fit the activation and deactivation kinetics for KCa3.1 L281W. (A) Representative activation and deactivation records fit with the model shown in Fig. 10 A. Activation and deactivation currents were recorded and fit according to the protocol described in the Materials and methods sections. The colors refer to the different Ca²⁺ concentrations, with red representing 0.5, blue 0.7, green 1.0, and black 10 µM Ca²⁺. The dashed line represents the fit assuming Ca²⁺-dependent rate constants have a nonlinear dependence on Ca²⁺ concentration, k = A · [Ca]/(B + [Ca]). Rate constants derived from the fit are summarized in Table II. (B) Enhanced depiction of the fit outlined by the box in A.

Figure 15.

Homology model of the pore region for the open and closed state of KCa3.1. (Left) Representation of the S5-P-helix-S6 region viewed as a profile illustrating the predicted orientation of Cys²⁷⁶ (yellow) and the L281W mutation (red) in the closed state using the KcsA structure. (Inset) Structure of PCMBS adapted from the crystal structure of PCMBS in complex with the Charcot-Leyden crystal protein (Ackerman et al., 2002). PCMBS is 6.5 Å in length and 9.0 Å in length when the thiol group is included. The color red represents oxygen molecules, yellow (sulfur), cyan (carbon), and white (mercury). (Right) Representation of the S5-P-helix-S6 region viewed as a profile illustrating the predicted orientation of Cys²⁷⁶ (yellow) and the L281W mutation (red) in the open state based upon the structure of rKv1.2. Sequence alignments used for homology modeling can be found in Fig. 5 C. (Inset) Representation of S5 and S6 viewed from the intercellular aspect of the pore to illustrate that Leu²⁸¹ (L281W) is oriented toward S5. Cys²⁷⁶ and L281W from the adjacent subunit are the only residues illustrated to compare the predicted orientation in the closed and open state. The S5 and P-helix are colored green and the S6 helix is colored blue. All residues are illustrated in VDW format. The C-terminal portion of the S5 helix was removed to prevent obscuring the view of L281W.