Role of the activation gate in determining the extracellular potassium dependency of block of HERG by trapped drugs

Abstract

Drug induced long QT syndrome (diLQTS) results primarily from block of the cardiac potassium channel HERG (human-ether-a-go-go related gene). In some cases long QT syndrome can result in the lethal arrhythmia torsade de pointes, an arrhythmia characterized by a rapid heart rate and severely compromised cardiac output. Many patients requiring medication present with serum potassium abnormalities due to a variety of conditions including gastrointestinal dysfunction, renal and endocrine disorders, diuretic use, and aging. Extracellular potassium influences HERG channel inactivation and can alter block of HERG by some drugs. However, block of HERG by a number of drugs is not sensitive to extracellular potassium. In this study, we show that block of WT HERG by bepridil and terfenadine, two drugs previously shown to be trapped inside the HERG channel after the channel closes, is insensitive to extracellular potassium over the range of 0 mM to 20 mM. We also show that bepridil block of the HERG mutant D540K, a mutant channel that is unable to trap drugs, is dependent on extracellular potassium, correlates with the permeant ion, and is independent of HERG inactivation. These results suggest that the lack of extracellular potassium dependency of block of HERG by some drugs may in part be related to the ability of these drugs to be trapped inside the channel after the channel closes.

Keywords: HERG; drug blockade; drug trapping; drug-induced long QT syndrome; extracellular cations; extracellular potassium; knockoff.

Figure 1. Block of WT HERG by bepridil and terfenadine does not depend on extracellular potassium. Summary of a number of different experiments of bepridil and terfenadine block of WT HERG in the presence of low (0K) and high (20K) extracellular potassium. Error bars indicate standard error of the mean; n values are as follows: bepridil: 0K(5),20K(3); terfenadine: 0K(3),20K(5). The voltage protocol shown in the figure was repeated every 6 sec and blockade assessed as described in the materials and methods section and in more detail previously. Traces above histogram: the inset above each histogram shows a representative block experiment from one oocyte and shows the current assessed at the beginning of the pulse to -60 mV plotted as a function of time for each repetitive pulse. After a steady-state current level was reached in the absence of drug, either 1 μM bepridil or 1 μM terfenadine was perfused into the bath. The vertical arrow and larger diamond indicates the time point when perfusion of drug into the bath was initiated. Data were sampled at 500 μsec and filtered at 1 kHz.

Figure 2. Block of D540K by bepridil and terfenadine depends on extracellular potassium. Summary of bepridil and terfenadine block of the HERG mutant D540K in the presence of low and high extracellular potassium. Error bars indicate standard error of the mean; n values are as follows: bepridil: 0K(6),20K(3); terfenadine: 0K(3),20K(3). An asterisk indicates a statistically significant difference between low and high potassium solutions (p < 0.05). The voltage protocol shown in the figure was repeated every 6 sec and blockade assessed as described in the materials and methods section. Traces above histogram: the inset above each histogram shows a representative block experiment from one oocyte and shows the current assessed at the beginning of the pulse to -60 mV plotted as a function of time for each repetitive pulse. After a steady-state current level was reached in the absence of drug, either 1 μM bepridil or 1 μM terfenadine was perfused into the bath. The vertical arrow and larger diamond indicates the time point when perfusion of drug into the bath was initiated. Data were sampled at 500 μsec and filtered at 1 kHz.

Figure 3. Bepridil dose response of WT vs. D540K in low and high potassium. Data were collected as described in Figures 1 and 2, and are plotted as the average of between 3 and 5 oocytes for each drug concentration. In some cases more than one concentration of drug was obtained in the same oocyte. The voltage protocol, shown next to each plot was repeated every 6 sec and blockade assessed as described in the materials and methods section. Solid line (0K), dashed line (20K). Error bars indicate standard error of the mean. IC₅₀ was estimated by fitting the data to 1/[1+(IC₅₀ⁿ/[drug] ⁿ)], where [drug] is the drug concentration and n is the Hill coefficient. Estimated IC₅₀ and hill coefficient: WT: 0Κ (0.69 μM, 0.94); 20K (0.92 μM, 1.05); D540K - 0K (0.75 μM, 1.15); 20K (1.58 μM, 1.69).

Figure 4. Estimation of WT and D540K permeability. For both D540K and WT, oocytes were depolarized to +30 mV for 2 sec, hyperpolarized to -160 mV for 10 msec, then depolarized to voltages between -150 and +50 mV for 400 msec. Oocytes containing WT were held at -100 mV and oocytes containing D540K were held at -80 mV. The current at the beginning of the last pulse was plotted as a function of voltage and the reversal potential was obtained by a linear regression of the 3–4 points near where the current reversed sign. D540K Dep, depolarization-activated D540K current; D540K Hyp, hyperpolarization-activated D540K current. Current was sampled at either 10 or 200 μsec and filtered at 1 kHz. Error bars represent standard error of the mean. n ranges from 3 to 13 oocytes depending on the solution.

Figure 5. Bepridil block of D540K in the presence of extracellular cations. Summary of block of D540K by 1 μM bepridil in the following extracellular solutions: 0K,20K,20NH₄,20Cs, and 40TEA. Data were collected as in Figures 1 and 2. The voltage protocol shown in the figure was repeated every 6 sec and blockade assessed as described in the materials and methods section. Error bars indicate standard error of the mean; n values are as follows bepridil 0K(6), 20K(3), 20NH₄(4),20Cs(6),40TEA(4) An asterisk indicates a statistically significant difference between the 0K solution and solution X (X = 20K, 20NH₄,20Cs, 40TEA) (p < 0.05).

Figure 6. Bepridil dose response for D540K in different extracellular cations. Data were collected as described in Figures 1 and 2, and are plotted as the average of between 3 and 7 oocytes for each drug concentration. In some cases more than one concentration of drug was obtained in the same oocyte. The voltage protocol shown in the figure was repeated every 6 sec and blockade assessed as described in the materials and methods section. Interrupted dashed line (40 TEA); solid line (0K), dotted line (20 Cs), dashed line (20K). Error bars indicate standard error of the mean. IC₅₀ was estimated by fitting the data to 1/[1+(IC₅₀ⁿ/[drug] ⁿ)], where [drug] is the drug concentration and n is the Hill coefficient. Estimated IC₅₀ and hill coefficient: WT: 40ΤΕΑ (0.93 μM, 0.90), 0Κ (0.75 μM, 1.15); 20Cs (1.32 μM, 1.41); 20K (1.58 μM, 1.69)

Figure 7. Estimation of HERG inactivation. For both (A and B) oocytes were held at -100 mV, depolarized to +20 mV for 1 sec, hyperpolarized to -100 mV for 25 msec, then depolarized to +20 mV for 400 msec. Current during the second pulse to +20 mV was fit to a single exponential function and the current extrapolated back to the beginning of the pulse to +20 mV. (A) Fraction of inactive channels: the fraction of inactive channels was calculated using a slight modification of the method described in the literature as 1–I_e/I_p where I_e is the current extrapolated back to the beginning of the second pulse to +20 mV and I_p is the steady-state current at the end of the first pulse to +20 mV. An identical protocol was used with depolarizing pulses to +10, 0 and -10 mV. As long as the two depolarizing pulses in the protocol are the same, the fraction of inactive channels is 1–I_e/I_p. In most cases current was sampled at 200 μsec and filtered at 1 kHz. In some cases the current was sampled at 10 μsec. Error bars represent standard error of the mean; n ranges from 3–8 oocytes for each solution. (B) Inactivation time constant: The current resulting from the 2nd depolarizing pulse in the protocol described in (A) was fitted with a single exponential function.

Figure 8. Bepridil block of D540K at -120 mV. Oocytes were held at -120 mV and current monitored every 6 sec. After a steady-state current level was reached in the absence of drug, a series of increasing concentrations of bepridil was perfused into the bath. Figure 10A shows block in 0K after addition of 3, 10, 50 and 100 μM and Figure 10B shows block in 20K after addition of 10 and 50 μM. The vertical arrow and larger diamond indicates the time point when perfusion of drug into the bath was initiated. Data were sampled at 500 μsec and filtered at 1 kHz. Insets: Each inset shows two sets of raw data traces: (1) inset on the left: 2 sec pulses to -160, -120 and -80 mV to show the time course of D540K current at negative potentials. (2) Inset on the right: raw steady-state currents before and after drug addition. Each time course and set of two insets represents one representative oocyte for that solution. P.D., pre-drug.

Figure 9. Bepridil dose response for D540K at -120 mV in different extracellular cations. Data were collected as described in Figure 8, and are plotted as the average of between 3 and 7 oocytes for each drug concentration. In some cases more than one concentration of drug was obtained in the same oocyte. Interrupted dashed line (20 NH₄); solid line (0K), dotted line (20 Cs), dashed line (20K). Error bars indicate standard error of the mean. IC₅₀ was estimated by fitting the data to 1/[1+(IC₅₀ⁿ/[drug] ⁿ)], where [drug] is the drug concentration and n is the Hill coefficient. Estimated IC₅₀ and hill coefficient: WT: 20 ΝΗ₄ (10.30 μM, 1.19), 0Κ (11.19 μM, 1.61); 20Cs (44.12 μM, 1.26); 20K (32.79 μM, 1.56).

Figure 10. Time course into and recovery from 0K solution. (A) After extracellular solution change from ND96 to 5K, oocytes containing either WT HERG or D540K were repetitively pulsed as described in Figures 1 and 2 and the Materials and Methods and the current level was assessed at the beginning of the pulse to -60 mV. After a steady-state current level was reached in 5K, either 0K, 20K or 20NH₄ (solution X) was perfused into the bath. (A) shows the time constant to reach steady-state for WT and D540K in either 0K, 20K or 20NH₄. (B) shows the time constant for recovery from either 0K, 20K or 20NH₄ (solution X) back to 5K.

Figure 11. Drug block in 0K. (Α) Data were collected and analyzed as described in Figure 1 and are plotted as the average of between 3 and 6 oocytes. Comparison of block of WT by 1 μM bepridil in 0K, 20K, and 20NH₄. (B) Data were collected and analyzed as described in Figure 8 and are plotted as the average of between 3 and 6 oocytes. Comparison of block of D540K by 3 μM bepridil in both 0 mM extracellular K and 20 mM extracellular K with a holding potential of either -120 or -80 mV.