A generic binding pocket for small molecule IKs activators at the extracellular inter-subunit interface of KCNQ1 and KCNE1 channel complexes

The cardiac IKs ion channel comprises KCNQ1, calmodulin, and KCNE1 in a dodecameric complex which provides a repolarizing current reserve at higher heart rates and protects from arrhythmia syndromes that cause fainting and sudden death. Pharmacological activators of IKs are therefore of interest both scientifically and therapeutically for treatment of IKs loss-of-function disorders. One group of chemical activators are only active in the presence of the accessory KCNE1 subunit and here we investigate this phenomenon using molecular modeling techniques and mutagenesis scanning in mammalian cells. A generalized activator binding pocket is formed extracellularly by KCNE1, the domain-swapped S1 helices of one KCNQ1 subunit and the pore/turret region made up of two other KCNQ1 subunits. A few residues, including K41, A44 and Y46 in KCNE1, W323 in the KCNQ1 pore, and Y148 in the KCNQ1 S1 domain, appear critical for the binding of structurally diverse molecules, but in addition, molecular modeling studies suggest that induced fit by structurally different molecules underlies the generalized nature of the binding pocket. Activation of IKs is enhanced by stabilization of the KCNQ1-S1/KCNE1/pore complex, which ultimately slows deactivation of the current, and promotes outward current summation at higher pulse rates. Our results provide a mechanistic explanation of enhanced IKs currents by these activator compounds and provide a map for future design of more potent therapeutically useful molecules.


Introduction 8 5
Potassium ion (K + ) channel activators are important compounds in human health, as partial or 8 6 complete loss of function of many K + channels may lead to inherited or acquired diseases that  helices of two separate subunits, and above by pore domain residues (Willegems et al., 2022). The location of the binding pocket and its structural inter-relationships help to explain the 1 0 8 underlying mechanism of action of ML277, its specificity for KCNQ1, and the lack of efficacy 1 0 9 due to steric hindrance in the presence of a β -subunit. However, it is known neither where isothiocyanatostilbene-2,2'-disulfonic acid), and diclofenac acid derivatives such as mefenamic 1 1 3 acid (Abitbol et al., 1999;Zheng et al., 2012;Wang et al., 2020), nor how they mediate their 1 1 4 activator action. We have some clues for DIDS and mefenamic acid that their binding sites are 1 1 5 not in the central channel core, as is the case for ML277, but depend on KCNE1 β -subunit 1 1 6 residues at the extracellular surface of the channel, residues 39-43 in KCNE1 for DIDS (Abitbol that residues in either of these regions could also provide important clues to explain mefenamic 1 2 4 acid's mechanism of action. To further explore the dependence of residues in KCNE1 and those in adjacent KCNQ1 sites on 1 2 7 binding of mefenamic acid to I Ks , we first examined the role of K41C in preventing the drug binding to I Ks followed by mutagenesis were used to map out critical KCNE1 and KCNQ1 1 3 0 residues. Further, we expanded on the idea that the stilbene, DIDS (Abitbol et al., 1999), which 1 3 1 is structurally quite different from mefenamic acid, shares a common binding site. Our results 1 3 2 showed that both compounds bind in the same general region formed by elements of the pore and critical residues for their binding stability. Exposure of the channel complex to either compound 1 3 5 induces subtle structural changes that subsequently stabilize the conformation of the S1/outer 1 3 6 pore/S6 of KCNQ1 and slow the I Ks deactivation gating kinetics. The results suggest the 1 3 7 existence of a common drug-induced binding site and a mechanism of action for small molecule 1 3 8 I Ks activators which is distinct from that of specific compounds that activate KCNQ1 alone. Calculations of free interaction energy shown in Table 6  in one simulation run, showed significantly lower free energy than DIDS binding to WT, using 3 7 5 both methodologies. The small change in free energy and more stable binding of DIDS to K41C 3 7 6 suggest that this residue is not as important as W323 or Y46 for DIDS binding. In longer 500 ns  both W323A and Y46C mutants unbound during 500 ns simulations (all runs for Y46C, Figure 6 3 8 2 -figure supplement 3) as was seen with the with AMBER force field. In this model the major K41C-EQ G-V plots were obtained from peak tail current amplitudes in the absence and µM DIDS hyperpolarized the V 1/2 and changed the shape of the G-V relationship compared to 3 9 7 control ( Figure 7B, 7D, Table 5). The current waveform of K41C-EQ activated more quickly 3 9 8 with less sigmoidicity after treatment with DIDS, and tail current decay was slowed as indicated 3 9 9 by the normalized response ( Figure 7B, 7C). As K41C-EQ remained responsive to DIDS, and 4 0 0 L42C responded similarly to K41C-EQ (for K41C-EQ, Δ V 1/2 was -24.7 mV, and for L42C-EQ shift in V 1/2 was also reduced compared to WT-EQ (A44C-EQ Δ V 1/2 = -18.5 mV; WT EQ Δ V 1/2 4 0 5 = -46.6mV, Figure 7C, D). In Y46A, the effect of DIDS on the current waveform was greatly 4 0 6 reduced as was slowing of tail current decay ( Figure 7A, 7C). We could only estimate an 4 0 7 activation V 1/2 for Y46A as the G-V curve was right shifted to very positive potentials and non- in the presence of DIDS was observed, suggesting that unstable and short-lived binding of the 4 1 0 drug to the channel complex was sufficient to interfere with channel gating. The modeling studies ( Figure 6) suggested that W323 remained a key residue for I Ks activator 4 1 3 sensitivity, and in agreement with this, EQ-W323A was only partially responsive to DIDS. In EQ-W323A, the tail current decay was affected but the slowly-activating current waveform was complex between pulses ( Figure 7A, 7C, Table 6). However, a significant shift in the V 1/2 4 1 7 remained (ΔV 1/2 = -27.3 mV; Figure 7D, segment. In EQ-Y148C, the V 1/2 and the shape of the G-V plot were not altered by 100 µM 4 2 1 DIDS ( Figure 7D, Table 5), and tail current decay was not markedly slowed ( Figure 7A, 7C).

2 2
This result suggests that unlike mefenamic acid, DIDS binding determinants extend to the S1 4 2 3 region of KCNQ1 and that Y148 is an important contributor.
The extracellular region of KCNE1 between residues 39-44 was found important in mediating 4 3 7 the effect of DIDS on I Ks (Abitbol et al., 1999) and residue K41, located on the extracellular end between the immediate extracellular residues of KCNE1, S1 and pore residues of two KCNQ1 mutations that negate the activator compound actions is to destabilize the binding pocket itself, 4 5 0 reduce drug binding and limit activator residency time on the channel complex. Using data that suggested an extracellular binding site for mefenamic acid was at the KCNE1- swapped S1 helices of one KCNQ1 subunit and the pore/turret region made up of two other particularly loss of interaction strength with KCNE1 (Table 7). Likewise, MD simulations  with a reduced response in terms of changes in waveform, V 1/2 and slope. In contrast, while the during 300 ns MD simulations was not observed (Supplemental movie S4, when placed in a lipid environment mefenamic did completely unbind in two out of five 500 ns Q147 and Y148 in S1 with this mutant and a small shift towards the turret at S298 and A300 4 8 2 ( Figure 8A; Table 7). While we were not able to obtain good electrophysiological data from that 100 µM DIDS enhanced WT EQ activity ( Figure 5B) with a V 1/2 shift of -46.6 mV and a 4 9 2 decrease in the slope of the G-V curve ( found on the surface of oocytes, higher drug concentrations than those used for cultured cells are transfected tsA201 cells were used. The results of in silico experiments, including some binding properties and stability of some complex (compare Movies S1 and S6). DIDS associated more strongly with the pore of the 5 1 0 Y148C mutant, particularly with V324 and T327, and less across the ps-KCNE1, S1 and turret 5 1 1 regions ( Table 7). The Y46C mutation resulted in shifts away from ps-KCNE1 and towards the 5 1 2 S1 domain (T144, E146 and Q147). In the case of A44C and DIDS, the changes were more and S1 (V141 and L142; coloured grey in Figure 8B) to more peripheral residues (I145, E146 5 1 5 and Q147) and lower in the KCNE TMD (A44 and M45; coloured magenta in Figure 8B). Proposed mechanism of action for mefenamic acid and DIDS. changes in the channel upon binding, to shape a binding pocket formed by residues from the 5 2 2 external S1 domain, KCNE1 and the pore domain of I Ks (Figure 2, Supplemental movies S1, S5).

2 3
This cryptic binding pocket is not detectable in the absence of the drug ( Figure 4B), which 5 2 4 suggests that it has been induced in a similar manner to previous reports of toxin interactions 5 2 5 with KcsA-Kv1.3 that induce conformational changes in both the toxin and the channel structure to generate a high-affinity binding site (Lange et al., 2006;Zachariae et al., 2008). Indeed, analysis of the binding site before and after mefenamic acid unbinds, shows the involved figure supplement 2), a collapse of the pocket created by the drug-channel interactions. where mefenamic acid and DIDS bind/unbind from I Ks at high frequency. Slowed deactivation 5 3 5 may therefore be the result of these rapid binding/unbinding interactions slowing the dissociation 5 3 6 of the S1/KCNE1/pore domain/drug complex by either providing steric hindrance to dissociation or by stabilizing the activated complex. MD simulations suggest the latter is most likely the case.

3 8
Crosslinking studies have previously shown that placing cysteines at key locations in the KCNE1 5 3 9 N-terminus, the top of S1 and in S6 can lead to disulfide bond formation and slowing or would not only destabilize the external S1/KCNE1/pore domain interface ( Figure 4D, Table 1) but also eliminate direct hydrophobic contacts which normally occur between the W323 side

4 6
A reduced interaction with S1 could also conceivably curtail the ability of I Ks activators to slow 5 4 7 dissociation of the activated complex and restore faster deactivation rates such that there is no longer an enhanced step current in the presence of the drug, as seen in mutants K41C and A44C importance of this S1-pore coupling to channel function is supported by mutational analyses of mutation can mirror the effects of the two activators studied here. Binding of mefenamic acid and DIDS to the extracellular end of KCNE1 and the KCNQ1 S6 and 5 9 2 S1 helices is facilitated by a number of key residues. Residue K41 acts as a "lid" holding mutated, as in the case of K41C-EQ, EQ-W323A and A44C-EQ, little to no effect of the drug is 5 9 7 seen. The larger drug, DIDS, interacts with many of the same residues but those deeper in the 5 9 8 pocket appear more important than for mefenamic acid. Furthermore, the qualitative similarities 5 9 9 between the S1 mutant channel, EQ-L142C and WT EQ in the presence of 100 µM mefenamic 6 0 0 acid suggest that I Ks activators most likely cause their effects by modulating interactions between 6 0 1 the S1 helix, pore turret, KCNE1 and the S6 helix. Upon binding, both DIDS and mefenamic Y58, were substituted with homologous KCNE1 residues D39-A44. Conformational sampling 6 1 7 was then performed for substituted residues and the lowest free energy conformations were The docking and conformational sampling as well as substitution of amino acids were performed 6 2 7 with ICM-pro 3.8 software (Neves et al., 2012).
The coordinates of mefenamic acid-bound ps-I Ks channel complexes with the lowest free energy 6 2 9 were then used for two sets of MD simulations with AMBER and CHARMM force fields (see of 1000 kJ/mol/nm 2 during MD simulations. Three 300 ns duration simulations were performed in a water environment with AMBER20 6 3 6 using a ff14SB force field for protein and GAFF/AM1-BCC scheme for the ligand 100 mM concentration. The system was minimized and equilibrated in the NVT and NPT 6 4 0 ensembles for 10 ns, gradually releasing spatial restraints from the backbone and sidechains. Five 500 ns duration simulations were performed using a CHARMM 36m force field and ps-I Ks  Trajectories from MD simulations were clustered with TTclust based on ligand and its binding RMSF calculations only backbone and C-beta atoms were used.

5 9
The 2D diagrams and other molecular visualizations were generated by ICM-pro and VMD Conductance-voltage (G-V) relationships were obtained from the normalized peak of the initial 6 8 9 tail current (G/Gmax) and plotted against the corresponding voltage. G-V plots were fitted with a 6 9 0 Boltzmann sigmoid equation to obtain the voltage at half-maximal activation (V 1/2 ) and slope (k) presence of drug-V 1/2 control) was also determined ( Figures 3E and 7D). In some cases, the foot DIDS impact on WT and mutated EQ channel complexes. Specifically, the peak to end 6 9 7 difference currents were calculated by subtracting the minimum amplitude of the deactivating 6 9 8 current from the peak amplitude of the deactivating current. The difference current in mefenamic 6 9 9 acid or DIDS was normalized to the maximum control (in the absence of drug) difference current 7 0 0 and subtracted from 1.0 to obtain the normalized response ( Figures 3D and 7C). applicable, unpaired t-test or one-way ANOVA followed by the Fisher's least significant Kasuya G, and Nakajo K (2022) Optimized tight binding between the S1 segment and KCNE3 is  sensor and pore are required for the function of voltage-dependent K(+) channels. PLoS.Biol. 7:e47. Lett. 22:5936-5941. Wu X, Brooks BR, and Vanden-Eijnden E (2016) Self-guided Langevin dynamics via  progressive continuum of pharmacological sensitivity by KCNQ potassium channels.            Figure depicts the binding pocket viewed from above (left) and the side (right) during a 1250 ns MD simulation before drug detachment (grey superimposed structure) and after (red superimposed structure). Mefenamic acid is represented as licorice and colored green. Snapshots for superimposition were collected every 10 ns. Visible from the snapshots is that when the drug leaves the binding site (after 500 ns, structures colored red), the N-terminal residues of ps-KCNE1, as well as W323 and other residues that form the pocket, shift toward the binding site that mefenamic acid previously occupied.

Top View
Side View WT EQ current in control (black) and exposed to 100 µM DIDS over time (grey). Pulses were from -80 to +60 mV every 15 s, and current traces are shown superimposed. Lower panel shows no effect on currents from GFP-only transfected cell exposed to 100 µM DIDS over time (grey). (C) Current traces from WT EQ in control and presence of 100 µM DIDS as indicated. Pulses were from -80 to +100 mV for 4s, with a 1 s repolarization to -40 mV. Interpulse interval was 15 s.  Size of residue ellipse is proportional to the strength of the contact. Light grey indicates residues in van der Waals contacts, light green hydrophobic contacts, and light blue are hydrogen bond acceptors. Red borders indicate KCNE1 residues, yellow are KCNQ1 VSD residues, and blue are pore residues. Dashed lines indicate H-bonds. The distance between the residue label and ligand represents proximity. Grey parabolas represent accessible surface for large areas. The 2D diagram was generated by ICM pro software with a cut-off value for hydrophobic contacts 4.5 Å and hydrogen bond strength 0.8. (C) Energy decomposition per amino acid for DIDS binding to ps-I Ks . Generalized Born Surface Area (MM/GBSA; orange) and Poisson-Boltzmann Surface Area (MM/PBSA; blue) methods were used to estimate the interaction free energy contribution of each residue in the DIDS-bound ps-I Ks complex. The lowest interaction free energy for residues in ps-KCNE1 and selected KCNQ1 domains are shown as enlarged panels (n=3 for each point). Error bars indicate ±SD.  Voltage steps from a holding potential of -80 mV to +70 mV for 4 s, followed by repolarization to -40 mV for 1 s. Interpulse interval was 15 s. Error bars shown are ± SEM. All calibration bars denote 0.5 nA/0.5 s.
(C) Summary plot of the normalized response to 100 µM DIDS. Data are shown as mean ± SEM and *p<0.05, **p<0.01, ***p<0.001 denote significant change in mutant versus WT currents (one-way ANOVA, see Materials and Methods). For calculation, see Materials and Methods. (D) Change in V 1/2 (ΔV 1/2 ) for WT EQ and each I Ks mutant in control versus DIDS. Data are shown as mean ± SEM and unpaired t-test was used, where *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 indicate significant change in V 1/2 comparing control to the presence of drug. n-values for mutants in C and D are stated in Table 5. For binding site mutants, cells were held at -80 mV, then pulsed to -90 mV to +100 mV for 4 s followed by -40 mV for 2 s. G-V plots in control (black) and presence of 100 μM Mef (colors). Error bars are ± SEM. Refer to Table 5 for Boltzmann fit values.

Supplemental movies 1-4
Supplemental movies 1-4. MD simulations at the molecular level of binding of mefenamic Acid to WT EQ, and K41C-EQ, W323A-EQ, Y46C-EQ mutants. Note that videos are longer than the actual simulations, durations stated below.