Asymmetric Interplay Between K+ and Blocker and Atomistic Parameters From Physiological Experiments Quantify K+ Channel Blocker Release

Figures

• 

  FIGURE 1. Block of single-channel currents in KcvNTS in DPhPC membranes by cytosolic TPrA. (A,B) Representative time series measured in symmetrical 100 mM KCl at different membrane voltages between +160 mV and −160 mV (A) without TPrA and (B) with 0.1 mM TPrA. (C) Time series measured in different concentrations of TPrA at +120 and −120 mV. The scale bar refers to all three panels in this figure. C marks the closed state. The data for ±120 mV without TPrA are displayed in both (A,C). (D) Alignment of the pore loop of two viral K+ channels (KcvNTS, Rauh et al., 2017b), the one used in this study, and KcvPBCV–1 (Tayefeh et al., 2009), a bacterial K+ channel (KcsA, UniProtKB/Swiss-Prot: P0A334.1) and three eukaryotic potassium channels (Shaker, GenBank: CAA29917.1, hERG, GenBank: BAA37096.1, and human Kv1.5, NCBI Reference Sequence: NP_002225.2). Sequence conservation (∗identity, : conservative,. semi-conservative) was calculated with the Clustal Omega Webapp. The selectivity filter sequence is underlined, the pore helix is shaded in gray. (E) Schematic illustration of the blocking process, the binding site is only accessible from the cytosolic pore entrance.

• 

  FIGURE 2. Kinetics of the TPrA block. (A) Kinetic Markov model used for the fits of the amplitude histograms (Supplementary Figure S1). The states O (open) and F, M, S (three closed states of different dwell times) are those used previously in the 4-state model of gating in the unblocked KcvNTS (Rauh et al., 2017a, 2018). The blocked state B is introduced by the blocker TPrA. (B) Voltage dependence of the rate constant kBO=koffO measured at cytosolic TPrA concentrations of 0.05 mM ●, 0.1 mM , 0.5 mM , 1 mM , and 5 mM ο. (C) Voltage dependence of kOB=konO[TPrA] measured at the same concentrations as kBO, (D) Voltage dependence of konO as determined by dividing kOB by the concentration of TPrA in mol. Data points are the geometric mean of 3 to 4 individual channels, error bars represent the geometric standard deviation. Some data points around 0 mV are missing because of insufficient signal-to-noise ratio.

• 

  FIGURE 3. Relating the voltage dependence of the rate constant of blocker release in the open state (koffO) to the ion distribution at S4 adjacent to the binding site of the blocker. (A) 5-state model of ion hopping (Roux, 2005) with the rate constants kij (i, j = 1 to 5) obtained from the previous analysis (Rauh et al., 2018). Numerical values for the rate constants are given in Supplementary Table S1 (B,C). The probabilities Pm (m = 1 to 5) of the occurrence of the states 1 to 5 in the model of Roux (2005) calculated from Eqs. S4 to S9 for the case (B) that the channel is open and not blocked (redrawn from the original data of Rauh et al., 2018) and (C) that the channel is blocked (k12 = k21 = k32 = 0). P(S4) is the probability that there is an ion in binding site S4 (purple in panel A) close to the binding site of the blocker. According to panel (A), P(S4) = P3 + P4 + P5. The inset shows the probabilities for the open and blocked channel at 0 mV, as taken from panels (B,C). (D) Comparison of the measured koffO with the values predicted by Eq. 4 based on P(S4). Black line: Fit of the experimentally determined koffO (black circles, pooled from all TPrA concentrations) with 100 mM…

• 

  FIGURE 4. (A,B) K+ ion concentration profiles and (C,D) cumulative equilibrium constants as function of channel length coordinate z starting from the origin as shown in the inset in (A) from finite concentration (1 M, uv, solid lines) and infinite dilution (single hydrated ion, uu, dashed lines) 3D RISM calculations of KcsA (pdb code 2HVK) with (red) and without (purple) TBA placed in the cavity. (E,F) cumulative results for four reduced models with and without TBA comprising isolated carbonyl cages for KcsA (2HVK, Yohannan et al., 2007, crystalized with TBA and 1K4C (Zhou et al., 2001) crystalized without TBA), KirBac3.1 (3ZRS, Bavro et al., 2012) and the computationally determined filter model of KcvPBCV–1 (Tayefeh et al., 2009; Hoffgaard et al., 2015). (A,C,E) from calculations with original force field, (B,D,F) with down-scaled filter carbonyl charges leading to halved local dipole moments. The inset in panel (A) represents the full 2HVK system in cartoon representation. The inset in panel (F) shows a superposition of all reduced filter carbonyl groups in the same color code as the curves.