State-independent inhibition of the oncogenic Kv10.1 channel by desethylamiodarone, a comparison with amiodarone

Fig. 1. desAd inhibition of Kv10.1 channels. A K+ currents evoked by 250-ms pulses from −40 to + 60 mV applied in 10 mV steps, HP = −70 mV. Left panel, control IK; right panel, IK left after superfusing the cell with the extracellular solution containing 5 µM desAd. B Current (IK) vs. voltage (Vm) relationship of the traces in A, IK was assessed at pulse end. The lines are the least-squares fit of the points with Eq. 1 (see the “Methods” section) with parameters: control, Gmax = 47.6 nS; V1/2 = + 1.1 ± 1.5 mV, z = 2.2 ± 0.2; desAd, Gmax = 13.2 nS, V1/2 = − 12.9 ± 1.0 mV, z = 2.2 ± 0.2. Gmax varies from cell to cell depending on the (highly variable) level of channel expression. Notice the left shift of the voltage dependence of activation (ΔV1/2 ~ −14 mV). C Dose–response relationship at + 30 mV (see Text). The line is the least-squares fit of the data with a Michaelis–Menten equation (Eq. 2, Methods) with maximal fractional inhibition FImax = 0.81 and Kd = 1.2 µM, obtained from the linearized double-reciprocal plot. The inset shows the double-reciprocal plot of the data. The straight line is the least-squares fit of the points with parameters: Kd = 1.2 µM and 1/FImax = 1.39. The linear…

Fig. 2. desAd steady-state inhibition in an extended voltage range. A Representative IK recorded in the presence of 3.5 µM desAd. IK was evoked by a + 50-mV/300-ms pulse applied immediately after 1-s conditioning pulses of the indicated voltages, applied from the HP of −70 mV, as indicated. B Ipeak/Ipeak,max ratio at + 50 mV, as a function of the conditioning voltages. Note that peak IK increases from −140 to approximately − 100 mV (see Text and the “Discussion” section). C Points in B re-plotted for conditioning voltages > −100 mV, where IK decreases monotonically with voltage. The line through the points is the least-squares fit of the data with a Boltzmann equation plus an offset (Methods), with parameters: Ipeak,max = 0.8, z = 1.0, V1/2 = − 28 mV, offset = C = 0.31. Notice that desAd binds (inhibits) the channels even before they open; to best see this, the figure shows normalized chord conductance vs. Vm curves of control (solid line, parameters: z = 1.5 ± 0.1, V1/2 = + 1.1 ± 1.5 mV) and desAd blocked channels (dashed line, parameters: z = 2.2 ± 0.2, V1/2 = −12 ± 1.0 mV), obtained from four experiments as in Fig. 1A; [desAd] = 3.5 µM

Fig. 3. desAd dissociation from closed channels. A IK evoked by a + 60-mV/120-ms conditioning pulse applied from − 70 mV, in the presence of 3.5 µM desAd (black trace), at pulse end, the membrane was repolarized to − 140 mV for 5 ms, and thereafter, IK was elicited by a test + 60 mV pulse (red trace). Note that on the second pulse channels retained the level of block achieved at the end of the first pulse (see Text). This shows that desAd was trapped in the channels by the closing of the activation gate. B IK recorded as in A, except that the time spent at − 140 mV, between activating pulses, was 1 s. There was ~ 80% recovery from blockage. This shows that desAd dissociates from closed channels. C Average extent of desAd leakage from closed channels (IK recovery) after 5 vs. 1000 ms at −140 mV, n = 4

Fig. 4. Deactivation tail currents of desAd-blocked vs. control channels. A Tail IK at − 140 mV of control (black trace) vs. 3.5 µM desAd-blocked channels (blue trace). The red trace shows IK of blocked channels scaled to match control IK initial amplitude. B As in A but at the repolarization potential of −90 mV. Tail IK was recorded following a + 30/100-ms activation pulse, applied from the HP of −70 mV (not shown in the figure). C Average deactivation time constants. Time constants are not significantly different (p(− 140 mV) = 0.98, n = 5; p(− 90 mV) = 0.93, n = 5)

Fig. 5. A test for desAd inhibition of the Cole-Moore shift. A Control IK at + 30 mV evoked from either the HP of −140 or −70 mV, as indicated. Note the ~ single-exponential time course of IK evoked from −70 mV and the slower, sigmoidal, time course of IK evoked from −140 mV, characteristic of the Kv10.1 Cole-Moore shift. B IK illustrating the previously reported Ad inhibition of the Cole-Moore shift. The panel shows superimposed traces of control (red) vs. Ad (1 µM)-blocked channels, as indicated. See that Ad inhibits the lag and the markedly sigmoidal time course of control IK (see Text). C Superposed IK recorded as in B, under control (red) and 3.5 µM desAd conditions, see that desAd does not inhibit the Cole-Moore shift (see Text)

Fig. 6. desAd and Ad inhibit IK in an apparently non-competitive fashion. A Superposed traces illustrating IK block by 1 µM desAd. IK was evoked by a + 30-mV pulse applied from − 70 mV. B Traces illustrating IK block by 0.2 µM Ad, recorded as in A. C desAd (1 µM) plus Ad (0.2 µM) joint inhibition of IK. D Comparison of the average extent of inhibition exerted by desAd plus Ad (Test) vs. the expected level of inhibition considering either a mutually exclusive (competitive, labeled C) or a non-competitive (NC) model of drugs interaction (see the “Methods” section). E Competition plot (see Text): the % Ik inhibition at + 30 mV was assessed as a function of the proportion (molar fraction) of desAd and Ad, as the concentration of each of them causing ~ 60% inhibition was reciprocally changed as X as: Ax, Bx-1, with X from 0 to 1

Fig. 7. Qualitative IK simulations. A Control IK evoked by a + 50-mV pulse applied from either −70 or − 140 mV resting potentials, as indicated ([B] = 0, see Text). Note the Cole-Moore shift of IK evoked from −140 mV. B Control ([B] = 0) vs. desAd-blocked ([B] = 3.5 µM) IK at + 50 mV, both IK activated from −70 mV (channels dwell in Cnear at the start of the pulse). Note the single-exponential activation time course. C Control ([B] = 0) vs. desAd-blocked IK at + 50 mV, both IK activated from −140 mV (channels dwell in Cfar at the start of the pulse). Note the sigmoidal activation time course and the DesAd lack of inhibition of the Cole-Moore shift. D Comparison of desAd-blocked IK evoked from either −70 or −140 mV, as indicated. Note that peak IK is bigger, in an extent comparable to that seen in Fig. 2A, when channels are activated from −70 than when they are from −140 mV. This difference qualitatively explains the increasing phase of the steady-state inactivation curve. Kfast = 0.05 ms−1, kslow = 0.025 ms−1, Kr = 0.008 ms−1, Kb = 0.022 (ms*µM)−1, [B] = [desAd] = 3.5 µM