We have studied and compared the effects of bupivacaine with those induced by a new local anesthetic, IQB-9302, on human cardiac K+ channels hKv1.5, Kv2.1, Kv4.3, and HERG. Both drugs have a close chemical structure, only differing in their N-substituent (n-butyl and cyclopropylmethyl, for bupivacaine and IQB-9302, respectively). Both drugs blocked Kv2.1, Kv4.3, and HERG channels similarly. Bupivacaine inhibited these channels by 48.6 ± 3.4, 45.4 ± 12.4, and 43.1 ± 9.1%, respectively, and IQB-9302 by 48.1 ± 3.3, 36.1 ± 3.7, and 50.3 ± 6.6%, respectively. However, bupivacaine was 2.5 times more potent than IQB-9302 to block hKv1.5 channels (EC50 = 8.9 ± 1.4 versus 21.5 ± 4.7 μM). Both drugs induced a time- and voltage-dependent block of hKv1.5 and Kv2.1 channels. Block of Kv4.3 channels induced by either drug was time- and voltage-dependent at membrane potentials coinciding with the activation of the channels. IQB-9302 produced an instantaneous block of Kv4.3 and hKv1.5 channels at the beginning of the depolarizing pulse that can be interpreted as a drug interaction with a nonconducting state. Bupivacaine and IQB-9302 induced a similar degree of block of HERG channels and induced a steep voltage-dependent decrease of the relative current. These results suggest that 1) bupivacaine and IQB-9302 block the open state of hKv1.5, Kv2.1, Kv4.3, and HERG channels; and 2) small differences at the N-substituent of these drugs do not affect the drug-induced block of Kv2.1, Kv4.3, or HERG, but specifically modify block of hKv1.5 channels.
Local anesthetics block the generation and conduction of nerve impulses by inhibiting the current through voltage-gated Na+channels in the membrane of nerve cells (Hille, 1977; Hondeghem and Katzung, 1977; Strichartz, 1987). Bupivacaine is a potent amide local anesthetic widely used for long-lasting epidural anesthesia (Strichartz, 1987). However, bupivacaine-like local anesthetics are not selective Na+ channel blockers. In fact, at the same range of concentrations used in the clinical practice to block the generation and propagation of nerve action potentials, this type of local anesthetic also blocks K+ channels (Valenzuela et al., 1995a,b; Lipka et al., 1998). Moreover, at higher concentrations, they also block the L-type Ca2+channels, thus decreasing cardiac contractility and conduction velocity through the atrioventricular node (Strichartz, 1987; Sanchez-Chapula, 1988). Bupivacaine-induced block of K+ channels has been considered the molecular mechanism by which this drug induces a prolongation of the QTc interval of the ECG in anesthetized dogs (Kasten and Martin, 1985; Wheeler et al., 1988) and human volunteers receiving high doses of this anesthetic (Scott et al., 1989).
hKv1.5, Kv2.1, Kv4.3, and HERG channels are involved in the repolarization of the human cardiac action potential (Wang et al., 1993; Curran et al., 1995; Firek and Giles, 1995; Mays et al., 1995;Sanguinetti et al., 1995; Van Wagoner et al., 1997; Kaab et al., 1998). In fact, Kv1.5, Kv4.3, and HERG are considered to be the cloned counterparts of the I Kur,I TO, andI Kr, respectively (Fedida et al., 1993; Wang et al., 1993, 1999; Sanguinetti et al., 1995; Dixon et al., 1996; Feng et al., 1997). Moreover, the presence of mRNA encoding the expression of Kv2.1 channels as well as the Kv2.1 protein in human atria has been demonstrated, although there is not a direct relationship of this channel with a native potassium current (Van Wagoner et al., 1997). The pharmacological effects of bupivacaine on different K+ channels, includingHERG, Kv1.4, Kv4.3, and hKvLQT1+minK have been already studied after the expression of the corresponding cRNA inXenopus oocytes (Lipka et al., 1998). However, it has been reported that when studying the effects of lipophilic drugs (such as bupivacaine) that act from the inside of the cell membrane, the apparent drug potency is approximately 10 to 100 times lower when using whole oocyte recordings than cell-free patch recordings (free of yolk) (Yatani et al., 1993; Ficker et al., 1998). Therefore, it would be of interest to study the effects of bupivacaine on K+ channels expressed in mammalian cells, that exhibit closer pharmacological responses to human cardiac myocytes thanXenopus oocytes. IQB-9302 is a new amide type local anesthetic, chemically related to bupivacaine (Fig.1), synthesized as a less cardiotoxic alternative to bupivacaine and with similar potency to block Na+ channels (Gallego-Sandı́n et al., 1999; Ruiz-Nuño et al., 1999). In the present study, we have analyzed the electrophysiological effects of bupivacaine and IQB-9302 on hKv1.5, Kv2.1, Kv4.3, and HERG channels expressed in mammalian cell lines. Preliminary results have been published in abstract form (González et al., 2000).
Materials and Methods
Stably transfectedLtk − cells with the gene encoding the expression of hKv1.5 or Kv2.1 channels were cultured in Dulbecco's modified Eagle's medium with 10% horse serum and 0.25 mg/ml G418 (GIBCO, Paisley, Scotland, UK) in a 5% CO2 atmosphere as previously described (Valenzuela et al., 1995a, 1996). Chinese hamster ovary cultures were grown in Ham's F-12 medium with 10% fetal bovine serum and transiently transfected with the cDNA encoding the Kv4.3 orHERG channel (4 μg) and the cDNA encoding the CD8 antigen (0.5 μg) by use of lipofectAMINE. Before experimental use, cells were incubated with polystyrene microbeads precoated with anti-CD8 antibody (Dynabeads M450; Dynal, Oslo, Norway). Most of the cells beaded also had channel expression.
The intracellular pipette-filling solution contained 80 mM potassium aspartate, 50 mM KCl, 3 mM phosphocreatine, 10 mM KH2PO4, 3 mM MgATP, 10 mM HEPES-K, and 5 mM EGTA and was adjusted to pH 7.25 with KOH. The bath solution contained 130 mM NaCl, 4 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 10 mM HEPES-Na, and 10 mM glucose, and was adjusted to pH 7.40 with NaOH. IQB-9302 [1-(cyclopropylmethyl-2′,6′-pipecoloxylidide] (gift from Dr. A. Galiano, IQB-Inibsa S.A., Barcelona, Spain) and bupivacaine (Sigma Chemical Co., St. Louis, MO) were dissolved in distilled deionized water.
hKv1.5, Kv2.1, Kv4.3, and HERG currents were recorded at room temperature (20–22°C) using the whole-cell patch-clamp technique (Hamill et al., 1981) with an Axopatch 1C patch-clamp amplifier (Axon Instruments, Foster City, CA). Kv1.5, Kv2.1, and Kv4.3 currents were filtered at 2 kHz (four-pole Bessel filter) and sampled at 4 kHz; HERG currents were filtered at 100 Hz and sampled at 200 Hz. Micropipettes were pulled from borosilicate glass capillary tubes (GD-1; Narishige, Tokyo, Japan) on a programmable horizontal puller (Sutter Instrument Co., San Rafael, CA) and heat-polished with a microforge (Narishige). Micropipettes resistance were 1 to 3 MΩ. Maximum hKv1.5 current amplitudes at +60 mV averaged 2.2 ± 0.3 nA, mean uncompensated access resistance was 2.4 ± 0.5 MΩ, and cell capacitance was 11.2 ± 0.8 pF (n = 10). Thus, no significant voltage errors (<5 mV) were expected with the electrodes used.
In all cases, cells were held at −80 mV. After control data were obtained, bath perfusion was switched to drug-containing solution. The effects of drug infusion was monitored with test pulses to +60 mV or to +30 mV (for HERG currents), applied every 30 s until steady state was obtained (after ∼12 min). Steady-state current-voltage relationships (IV) were obtained by averaging the current over a small window at the end of 250-ms or 5-s (forHERG currents) depolarizing pulses. Between −80 and −40 mV only passive linear leak was observed and least-squares fits to these data were used for passive leak correction. Deactivating “tail” currents of hKv1.5 and Kv2.1 channels were recorded at −40 mV andHERG tail currents were recorded at −60 mV. The activation curve of hKv1.5, Kv2.1, and HERG channels were obtained from the tail current amplitude measured just after the capacitive transient. Block of Kv4.3 channels was calculated as the reduction of the amount of charge crossing the membrane during the application of 250-ms depolarizing pulses from −80 mV to +60 mV (estimated from the integral of the current). Command potentials, data acquisition, and measurements were done using the Clampfit program of pClamp 6.0.1, origin 5.0 (Microcal Software, Northampton, MA) and by a custom-made analysis program.
In the case of hKv1.5 channels, EC50 values were obtained from f = 1/[1 + (EC50/[D])nH], where [D] represents the drug concentration andn H, the Hill coefficient. The apparent rate constants for binding (k) and unbinding (l) were obtained from the following equation Equation 1where τB represents the time constant of the drug-induced fast initial decline during depolarization to +60 mV. Thus, k and l were calculated from the fit of 1/τB versus different drug concentrations. For Kv2.1 and Kv4.3 currents, τB values were obtained from the monoexponential fit of the (I Control −I Drug)/I Controlratio.
The dominant time constant of the activation process was analyzed fitting it to a single exponential, following a procedure previously described and used for the same purpose (Valenzuela et al., 1995a). Deactivation and inactivation were fitted to a biexponential process as follows: Equation 2where τ1 and τ2are the system time constants, A 1 andA 2 are the amplitudes of each component of the exponential, and C is the baseline value. Half-maximal voltages (E h) and slope factors (s) of activation of hKv1.5, Kv2.1, andHERG channels were determined by fitting data with a Boltzmann equation: y = 1/[1 + exp(−(E −E h)/s)]. The curve-fitting procedure used a nonlinear least-squares (Gauss-Newton) algorithm; results were displayed in linear and semilogarithmic format, together with the difference plot. Goodness of fit was judged by the χ2 criterion and by inspection for systematic nonrandom trends in the difference plot.
Voltage dependence of block was determined as follows: leak-corrected current in the presence of drug was normalized to matching control to yield the fractional block at each voltage (f = 1 −I Drug /IControl). The voltage dependence of block was fitted to the following equation: Equation 3where z, F, R, and Thave their usual meaning, δ represents the fractional electrical distance, i.e., the fraction of the transmembrane electrical field sensed by a single charge at the receptor site andK D* represents the apparent dissociation constant at the reference potential (0 mV).
Results are expressed as mean ± S.E.M. Direct comparisons between mean values in control conditions versus mean values in the presence of drug for a single variable were performed by a paired Student's t test. ANOVA was used to compare more than two groups. Student's t test was also used to compare two regression lines. Differences were considered significant if the p value was less than 0.05.
Effects of IQB-9302 and Bupivacaine on hKv1.5 Channels.
Figure2A shows current records obtained in the absence and in the presence of 20 μM bupivacaine or IQB-9302. The holding potential was maintained at −80 mV and 250-ms depolarizing pulses in duration to membrane potentials between −80 and +60 mV were applied. Tail currents were recorded upon repolarization to −40 mV. At 20 μM, both drugs inhibited hKv1.5 current, bupivacaine being more potent than IQB-9302 (57.3 ± 2.9%, n = 4 versus 41.6 ± 2.3%, n = 10, p < 0.01, respectively). The effects of both drugs were reversible upon perfusion of the cells with drug-free external solution (90 ± 3% of the control values, n = 6, and 82 ± 4%,n = 9, for bupivacaine and IQB-9302, respectively). Both anesthetics significantly accelerated the activation kinetics of the current at +60 mV (1.44 ± 0.08 versus 1.21 ± 0.09 ms,n = 5, p < 0.05; and 1.28 ± 0.11 versus 0.81 ± 0.01 ms, n = 4, p< 0.01, for bupivacaine and IQB-9302, respectively). The degree of inhibition of hKv1.5 current was measured at the end of depolarizing pulses from −80 to +60 mV induced by different bupivacaine and IQB-9302 concentrations (between 0.2 and 100 μM). Individual data were fit with the Hill equation to estimate an EC50 for hKv1.5 current inhibition and a Hill coefficient. These fits yielded EC50 values of 8.9 ± 1.4 μM (n = 22) and 21.8 ± 4.7 μM (n = 29, p < 0.01) for bupivacaine and IQB-9302, and Hill coefficients of 0.88 ± 0.02 and 0.92 ± 0.01, respectively.
Both drugs induced a fast initial decline of the current that was superimposed to the slow inactivation (Fig. 2A). This time-dependent decay of the current was more evident in the presence of bupivacaine than in the presence of IQB-9302, reflecting a different kinetics of block for each drug. The time constant of this drug-induced initial decline of the current was faster at higher concentrations and thus, it was considered as an index of the kinetics of binding of the drug (τB). Figure 3, A and B, show the apparent rate of block (1/τB) versus bupivacaine and IQB-9302 concentration. The straight lines are the least-squares fit to the relation 1/τB =k × [D] + l which for bupivacaine, yielded apparent association (k) and dissociation (l) rate constants of 2.45 ± 0.19 μM−1 s−1 and 29.1 ± 8.8 s−1 (n = 14), respectively. The k value for IQB-9302 was similar to that obtained for bupivacaine (2.07 ± 0.43 μM−1 s−1,n = 8, p > 0.05), whereas lwas ∼2 times faster than that obtained with bupivacaine (56.8 ± 9.2 s−1, n = 8,p < 0.01). These results suggest a more stable interaction between the receptor and bupivacaine than with IQB-9302 and explain differences in potency between both drugs. To further analyze the time course of development of block, in Fig. 3, A and B (insets), we represented the ratio between the drug-sensitive current and the current in control conditions [(I Control −I Drug)/I Control] during the first 100 ms in the presence of 20 μM bupivacaine or IQB-9302. In the presence of either drug, block developed following a monoexponential function with time constants of 11.3 ± 0.8 ms (n = 7) and 12.6 ± 1.4 ms (n = 7), respectively. Whereas bupivacaine development of block only appeared upon depolarization; an initial “instantaneous” component of block was observed in the presence of IQB-9302 at the beginning of the depolarizing pulse (t = 0 ms), which averaged 15.3 ± 2.0% (n = 7), and that was followed by the previously described time-dependent increase in block. In fact, IQB-9302 induced a higher degree of block measured at the maximum peak current than that produced by bupivacaine. Thus, block at the maximum peak current induced by bupivacaine and IQB-9302 contributed by 47.5 ± 4.2% (n = 4) and 77.4 ± 2.4% (n = 10, p < 0.01), respectively, to the total amount of block.
A time-dependent block was also observed in the deactivating process. Figure 2B shows superimposed current traces obtained under control conditions and in the presence of 20 μM bupivacaine or IQB-9302. In the absence of bupivacaine, deactivation process of hKv1.5 current exhibited a biexponential decay with a fast time constant (τf) of 22.8 ± 1.6 ms and a slow one (τs) that averaged 95.9 ± 22.6 ms. Bupivacaine eliminated the fast component of deactivation, so that deactivation became a monoexponential process with a time constant of 97.2 ± 19.2 ms (n = 5, p > 0.05 versus the slow time constant obtained under control conditions). IQB-9302 increased the two time constants from 19.0 ± 3.4 and 57.9 ± 15.0 ms to 28.2 ± 2.9 ms (n = 5,p < 0.05) and 112.5 ± 33.0 ms (n= 5, p > 0.05). Moreover, in the presence of both IQB-9302 or bupivacaine, an initial rising phase at the beginning of the tail current was observed, thus indicating the dissociation of the drug from the receptor at the channel. The slowing of the deactivating process induced by both drugs produced a “crossover” phenomenon when the tail currents obtained under control conditions and in the presence of the drug were superimposed, suggestive of an open channel block mechanism (Armstrong, 1971).
Figure 3C shows the IV relationships for hKv1.5 obtained in the absence and in the presence of 20 μM bupivacaine or IQB-9302. The degree of block induced by bupivacaine and IQB-9302 was higher at more positive membrane potentials, suggesting an open channel block mechanism. To quantitate this voltage dependence the relative current in the presence of bupivacaine or IQB-9302 was plotted versus the membrane potential (Fig. 3D). Block steeply increased in the activation range of hKv1.5 channels (dotted line) and it significantly increased in a shallower way at membrane potentials positive to 0 mV. This voltage dependence was explained as the consequence of the effects of the transmembrane electrical field on the interaction between the charged form of the drug and its receptor in the channel. Thus, following a Woodhull formalism, we calculated the δ-values, that averaged 0.20 ± 0.02 (n = 14) for bupivacaine. For IQB-9302, δ averaged 0.17 ± 0.01 (n = 10, p> 0.05).
Effects of IQB-9302 and Bupivacaine on Kv2.1 Channels.
Figure4A shows original records of Kv2.1 currents in the absence and in the presence of bupivacaine and IQB-9302. At 20 μM, bupivacaine and IQB-9302 induced a percentage of block of Kv2.1 channels that averaged 48.6 ± 3.4% (n = 6) and 48.1 ± 3.3% (n = 6,p > 0.05), respectively. Both drugs produced a decrease of the time constant of activation of the current [from 15.5 ± 2.3 to 11.7 ± 2.8 ms (n = 6,p < 0.01) in the absence and in the presence of bupivacaine, and from 22.5 ± 2.2 to 19.7 ± 2.8 ms (n = 6, p < 0.05), in the absence and in the presence of IQB-9302]. In contrast to the effects observed in hKv1.5 channels, neither drug induced a fast initial decline of the maximum activated Kv2.1 current, i.e., block did not display time dependence on the maximum activated outward current. However, as it can be observed in Fig. 4B, both drugs slowed the time course of deactivation of Kv2.1 channels. Under control conditions, the deactivation process was fitted following a biexponential process, with a fast and a slow time constant of 10.5 ± 0.7 and 54.2 ± 8.1 ms (n = 4), respectively. In the presence of bupivacaine, the deactivation process was monoexponential with a time constant of 37.1 ± 5.4 ms (n = 4,p > 0.05 compared with the control τs) and a crossover phenomenon was observed. IQB-9302 also slowed the deactivation process, but at the same concentration than bupivacaine this new local anesthetic slowed the fast time constant of deactivation (14.8 ± 1.4 versus 18.5 ± 1.2 ms, n = 5, p < 0.05), without modifying the slow one (77.7 ± 17.0 versus 80.0 ± 16.6 ms,n = 5, p > 0.05). To calculate the time constant of development of block, in Fig.5A we represented the ratio between the drug-sensitive current and the current in control conditions [(I Control −I Drug)/I Control]. The fit of this ratio to a monoexponential function yielded the time constant of development of block, which averaged 9.3 ± 0.6 ms (n = 5) and 11.7 ± 1.3 ms (n = 6,p > 0.05) for bupivacaine and IQB-9302, respectively.
Figure 5B shows the IV relationships for Kv2.1 obtained in the absence and in the presence of 20 μM bupivacaine or IQB-9302. Block of Kv2.1 channels induced by bupivacaine and IQB-9302 was voltage-dependent (Fig. 5C). Block steeply increased in the range of activation of Kv2.1 channels and it increased with a shallower slope at membrane potentials positive to +15 mV. Using the same procedure described above (Woodhull, 1973), we calculated the δ-values, which averaged 0.16 ± 0.01 (n = 5) for bupivacaine and 0.20 ± 0.03 (n = 4, p > 0.05) for IQB-9302.
Effects of IQB-9302 and Bupivacaine on Kv4.3 Channels.
Figure6A shows original records of Kv4.3 current after applying 250-ms depolarizing steps from a holding potential of −80 mV to membrane potentials between −80 and +60 mV in 10-mV steps in the absence and in the presence of 20 μM bupivacaine or IQB-9302. Both drugs decreased the peak current to a similar extent [25.3 ± 2.8% (n = 6) and 32.6 ± 3.0% (n = 7, p > 0.05), in the presence of bupivacaine and IQB-9302, respectively]. However, their most prominent effect was an acceleration of the time course of inactivation, which was faster in the presence of bupivacaine than in the presence of IQB-9302 (Fig. 6B). The fast time constants of inactivation decreased from 27.4 ± 3.9 to 9.9 ± 1.4 ms (n = 5,p < 0.01) in the presence of bupivacaine and from 26.8 ± 2.9 to 13.2 ± 1.3 ms (n = 5,p < 0.05) in the presence of IQB-9302, with this effect being reversible upon washout of the cells with drug-free external solution. This accelerated decline of the current was suggestive of an open channel block mechanism, indicating that the reduction of peak current would be a nonequilibrium measure of block. Therefore, block of Kv4.3 channels by bupivacaine and IQB-9302 was also measured as the reduction of the amount of charge crossing the membrane (estimated from the integral of the current) during the application of 250-ms depolarizing pulses from −80 to +60 mV. Bupivacaine and IQB-9302 produced similar levels of Kv4.3 inhibition, reducing the integrated current by 45.4 ± 12.4% (n = 6) and 36.1 ± 3.7% (n = 7, p > 0.05), respectively. Interestingly, the inhibition of the current induced by bupivacaine measured at the maximum peak current was lower than that obtained from the inhibition measurements of the integral of the current (25.3 ± 2.8 versus 45.4 ± 12.4%, n= 5, p < 0.05), whereas IQB-9302 inhibited similarly the maximum peak current than the current integral (32.6 ± 3.0 versus 36.1 ± 3.7%, n = 7, p > 0.05), suggesting that IQB-9302 may block other state of Kv4.3 channels previous to the open one.
Using the procedure previously described, the τB values for bupivacaine and IQB-9302 on Kv4.3 channels were calculated from the fitting of the ratio between the drug-sensitive current and the current in control conditions [(I Control −I Drug)/I Control] averaging 6.5 ± 0.7 ms (n = 4) and 7.5 ± 0.8 ms (n = 4, p > 0.05), respectively (Fig. 7A). As it is observed in Fig. 7A, although bupivacaine development of block began during the depolarization, IQB-9302 development of block exhibited an initial “instantaneous” component at the beginning of the depolarizing pulse (t = 0 ms), which averaged 19.2 ± 1.0% (n = 4) and that was followed by the previously described time-dependent increase in block.
Bupivacaine and IQB-9302 blocked Kv4.3 channels at all membrane potentials tested (Fig. 7B). However, and in contrast to that observed in hKv1.5 and Kv2.1 channels, block of Kv4.3 channels induced by bupivacaine or IQB-9302 was weakly voltage-dependent at membrane potentials positive to 0 mV, with δ-values measured from the inside of the membrane that averaged 0.11 ± 0.05 (n = 4) and 0.08 ± 0.04 (n = 4, p > 0.05) for bupivacaine and IQB-9302, respectively (Fig. 7C).
Effects of IQB-9302 and Bupivacaine on HERGChannels.
Figure 8A shows original records obtained after applying 5-s depolarizing steps from a holding potential of −80 mV to membrane potentials between −80 and +50 mV in 10-mV steps in the absence and in the presence of bupivacaine or IQB-9302 (20 μM). Under control conditions, outward HERGcurrents during depolarization are reduced because channels inactivate faster than they activate (Sanguinetti et al., 1995). Upon repolarization, rapid recovery from inactivation preceded deactivation, resulting in a hooked tail current (Sanguinetti et al., 1995). Bupivacaine and IQB-9302 decreased to a similar extent the amplitude of the current measured at 0 mV, 43.1 ± 9.1% (n = 5) and 50.3 ± 6.6% (n = 5, p > 0.05), respectively. Figure 8B shows tail HERG currents obtained in the absence and in the presence of 20 μM bupivacaine or IQB-9302. Deactivation kinetics of HERG channels at −60 mV after depolarization of the cell membrane to +50 mV exhibited a biexponential kinetics. Bupivacaine slowed the deactivation time course from 168.3 ± 22.8 and 845.7 ± 89.8 ms to 260.0 ± 22.5 ms (n = 4, p < 0.01) and 979.0 ± 104.7 ms (n = 4, p < 0.05), without modifying the contribution of the fast component to the total process (0.47 ± 0.06 versus 0.43 ± 0.03, in the absence and in the presence of bupivacaine, n = 4, p > 0.05). However, IQB-9302 did not modify the kinetics of the tail currents. Figure 9A shows the IV relationships obtained by plotting the magnitude of the HERGcurrent amplitude at the end of 5-s pulses as a function of the membrane potential under control conditions and in the presence of bupivacaine or IQB-9302 (20 μM). Both drugs inhibited to a similar extent HERG current at membrane potentials ranging between −20 and +60 mV. Figure 9B shows the activation curves obtained in the absence and in the presence of bupivacaine or IQB-9302. Both drugs shifted the activation curve toward hyperpolarizing direction without modifying its slope. In fact, the E hvalues in the absence and in the presence of bupivacaine averaged −3.1 ± 1.1 and −9.9 ± 1.7 mV, respectively (n = 4, p < 0.05). In the absence and in the presence of IQB-9302, E haveraged −8.0 ± 1.8 and −14.0 ± 2.3 mV (n= 5, p < 0.05). Figure 9C shows the relative tail current in the presence of bupivacaine or IQB-9302 versus membrane potential. The relative current steeply decreased in the membrane potential range coinciding with the HERG channels activation, suggesting an open channel block mechanism. Similarly to that observed in Kv4.3 channels, at membrane potentials positive to +15 mV, a slight increase in block was observed, consistent with δ-values of 0.09 ± 0.02 (n = 4) and 0.08 ± 0.04 (n = 6, p > 0.05) in the presence of bupivacaine and IQB-9302.
In the present article the effects of bupivacaine and IQB-9302 on hKv1.5, Kv2.1, Kv4.3, and HERG channels expressed in mammalian cells have been analyzed. Both drugs blocked Kv2.1, Kv4.3, and HERG channels to a similar extent, whereas bupivacaine resulted to be 2.5-fold more potent than IQB-9302 to block hKv1.5 channels. To our knowledge, this is the first study in which the effects of bupivacaine have been analyzed on hKv1.5 and Kv2.1 channels.
Effects of Bupivacaine and IQB-9302 on hKv1.5, Kv2.1, Kv4.3, andHERG Channels.
Block induced by bupivacaine and IQB-9302 of hKv1.5 channels was time- and voltage-dependent. Both drugs induced a fast initial decline of the current upon depolarization, which reached steady state at the end of 250-ms depolarizing pulses. Although bupivacaine block appeared only upon depolarization, suggesting a pure open channel block mechanism, IQB-9302 block showed an initial block previous to the development of block observed during depolarization, which suggests that this drug also blocks a closed state of the channel previous to the open one. The superposition of the tail currents recorded under control conditions and in the presence of either drug shows a “crossover” between them, indicating fast recovery from block during deactivation, consistent with an open channel block mechanism (Armstrong, 1971). Block induced by bupivacaine or IQB-9302 of hKv1.5 channels was voltage-dependent, in such a way that block steeply increased in the range of membrane potentials coinciding with the range of activation of hKv1.5 channels, suggesting that these drugs need that the channels open before they can bind to their receptor site and block K+ efflux. At membrane potentials positive to 0 mV, block induced by bupivacaine and IQB-9302 increased with a shallower slope, consistent with a δ-value of ∼0.17. This voltage-dependent block was interpreted to be the consequence of the effects of the transmembrane electrical field on the interaction between the cationic form of the drugs (pK a of ∼8.0) and their receptor site at the channel level, as it has been proposed for bupivacaine enantiomers in hKv1.5 channels (Valenzuela et al., 1995a).
Time-dependent block of Kv2.1 channels was evident in the deactivation process, but not at the maximum activated current during the depolarization step. Possible explanations for these results would be that bupivacaine and IQB-9302 bind to other state of Kv2.1 channels different from the open state and/or that they bind to the open state of Kv2.1 channels with a faster rate than the channel opening. The later explanation seems to be more likely. In fact, the time constants of block for bupivacaine and IQB-9302 (9 and 11 ms, respectively) were faster than the dominant time constant of activation of Kv2.1 current at +60 mV (15.5 ± 2.3 ms). As shown in Fig. 5A, both drugs induced the development of block during the activation of the channel, and this block increased exponentially until it reached steady-state block after ∼100 ms of the beginning of the depolarizing pulse. Moreover, block of Kv2.1 channels induced a slowing of the tail currents. These time-dependent effects may suggest an open channel block mechanism. In fact, block induced by bupivacaine and IQB-9302 of Kv2.1 channels was voltage-dependent consistent with a δ-value of ∼0.17. All these results suggest that bupivacaine and IQB-9302 block the open state of Kv2.1 channels.
Bupivacaine and IQB-9302 induced a faster inactivation process of the Kv4.3 current and block increased in an monoexponential manner during the activation of the channels, reaching a maximum value after ∼50 ms, consistent with an open channel block mechanism. However, in the presence of IQB-9302 an instantaneous block was observed att = 0 ms. This initial block, which appears before channel opening averaged 19.2 ± 1.0%, and could be attributed to the drug interaction with a nonconducting state of the channel. Consistent with this hypothesis, the degree of inhibition of the maximum peak current and the charge by IQB-9302 were similar, whereas for bupivacaine, both were significantly different, with the charge inhibition being higher than that of the peak current. However, as suggested by the further inhibition of the current observed during the application of depolarizing pulses, it seems that IQB-9302 also binds to the open state of the channel. Bupivacaine-induced block of Kv4.3 channels expressed in Xenopus oocytes exhibited an EC50 value of ∼33 μM (Lipka et al., 1998), close to that found in the present study, in which, if we assume ann H = 1, it would be ∼24 μM. Moreover, similarly to what has been previously described for bupivacaine on Kv4.3 channels expressed in Xenopus oocytes (Lipka et al., 1998), we found that the blockade produced by bupivacaine and IQB-9302 was weakly voltage-dependent at very positive membrane potentials, consistent with δ-values of ∼0.06.
HERG channels were also sensitive to the effects of bupivacaine and IQB-9302. The inhibition produced by 20 μM bupivacaine in the present study is similar to that reported forHERG channels expressed in Xenopus oocytes (100 μM) (Lipka et al., 1998). These differences found in both expression systems could be due to a different protein processing or membrane environment between amphibian oocytes and mammalian cells. In fact, differences between these two expression systems have been observed in pharmacological studies examining quinidine block of hKv1.5 and hKv1.4 channels (Yatani et al., 1993; Deal et al., 1996; Yeola and Snyders, 1997; Franqueza et al., 1999). Block of HERG channels induced by both drugs increased in the range of membrane potentials coinciding with the activation of the current, thus suggesting an open channel block mechanism. Similar to what was observed in Kv4.3 channels, block of HERG channels at positive membrane potentials was weakly voltage-dependent (δ of ∼0.07). Therefore, these results seem to suggest that both local anesthetics exhibit a higher affinity for the open state than for the inactivated state of the HERG channels.
Bupivacaine and IQB-9302 have the same chemical structure with the exception that at position 1, bupivacaine exhibits a butyl group and IQB-9302 a cyclopropylmethyl group (Fig. 1). Degree of block of Kv2.1, Kv4.3, and HERG channels induced by bupivacaine and IQB-9302 was very close, thus suggesting that both drugs act as nonspecific blockers of these K+ channels as previously found for bupivacaine in Xenopus oocytes (Lipka et al., 1998). However, in the present study, we found that the minor differences in the chemical structure between bupivacaine and IQB-9302 were sufficient to induce a 2.5-fold difference in potency to block hKv1.5 channels. Therefore, the present results suggest that, in contrast to the results obtained in Kv2.1, Kv4.3, and HERG channels, the affinity of the receptor site for local anesthetics at hKv1.5 channels is very sensitive to changes in the length of their N-substituent and/or to its torsion availability. Moreover, block of Kv2.1 and Kv4.3 channels by bupivacaine enantiomers is not stereoselective (Franqueza et al., 1997,1999), whereas block of hKv1.5 channels induced by bupivacaine and IQB-9302 is stereoselective (Valenzuela et al., 1995a; González et al., 2001). Since stereoselectivity indicates a direct and specific receptor-mediated action (Ariëns, 1993), these results may suggest that hKv1.5 channels are more sensitive to changes in the N-substituent than the other K+ channels studied.
Clinical Implications of the Present Study.
It has been described that bupivacaine decreases intracardiac conduction velocity and widens the QRS complex of the electrocardiogram, effects that were attributed to the inhibition of I Na(Clarkson and Hondeghem, 1985; Wheeler et al., 1988). Both in animal models and in humans, bupivacaine prolonged the duration of the cardiac action potential (Avery et al., 1984; Kasten and Martin, 1985; Scott et al., 1989; Solomon et al., 1990) and the QTc interval of the ECG (Scott et al., 1989), effects that can, eventually, result in the development of a polymorphic ventricular tachycardia known as torsades de pointes (Kasten and Martin, 1985). Accidental intravascular injection of bupivacaine can result in transient plasma concentrations of its free form of 4 to 12 μg/ml (12–36 μM) (Kotelko et al., 1984; Sage et al., 1985). Since the concentrations of bupivacaine used in the present study are within this range, it can be proposed that the prolongation of the QTc interval observed in patients who have received an overdose of bupivacaine can be explained by the blockade of several cardiac K+ channels (Scott et al., 1989). In the present experiments, IQB-9302 blocked human cloned potassium channels Kv2.1, Kv4.3, and HERG similarly to bupivacaine. Therefore, it would be expected that ventricular cardiotoxicity related to the blockade of K+ channels would be similar for both drugs. However, further clinical trials are needed to assess this possibility.
The results shown in the present article indicate that small differences at the N-substituent of bupivacaine-like local anesthetics do not affect the degree of block of Kv2.1, Kv4.3, orHERG channels, although specifically modify block of hKv1.5 channels, suggesting that hKv1.5 channels are more sensitive to changes in the N-substituent than Kv2.1, Kv4.3, or HERG channels.
We thank Guadalupe Pablo and Jose Luis Llorente for excellent technical assistance.
- Received July 31, 2000.
- Accepted October 30, 2000.
Send reprint requests to: Teresa González, B.S., Institute of Pharmacology and Toxicology, CSIC, School of Medicine, Universidad Complutense, 28040 Madrid, Spain. E-mail:
This study was supported by Comision Interministerial de Ciencia y Tecnologia SAF98-0058 (to C.V.), Comision Interministerial de Ciencia y Tecnologia SAF99-0069 (to J.T.), CAM 08.4/0016198 (to E.D.), and U.S.-Spain Science and Technology Program 98131 (to C.V.) Grants.
- ultrarapid delay rectifier potassium current
- transient outward potassium current
- rapid delay rectifier potassium current
- current-voltage relationship
- time constant of block
- fractional electrical distance
- voltage at which 50% of the channels are open
- The American Society for Pharmacology and Experimental Therapeutics