antiarrhythmic drugs in the cardiac Na+ channel

Transcription

1 Proc. Natl. cad. Sci. US Vol. 92, pp , December 1995 Pharmacology Molecular determinants of drug access to the receptor site for antiarrhythmic drugs in the cardiac Na+ channel (ion channels/pore blockers/local anesthetics/electrical excitability) YUSHNG QU, JOHN ROGRS, TIM TND, TODD SCHUR, ND WILLIM. CTIrRLL Department of Pharmacology, ox 35728, University of Washington, Seattle, W Contributed by William. Catterall, ugust 25, 1995 STRCT The clinical efficacy of local anesthetic and antiarrhythmic drugs is due to their voltage- and frequencydependent block of Na+ channels. Quaternary local anesthetic analogs such as QX-314, which are permanently charged and membrane-impermeant, effectively block cardiac Na+ channels when applied from either side of the membrane but block neuronal Na+ channels only from the intracellular side. This difference in extracellular access to QX-314 is retained when rat brain ril Na+ channel a subunits and rat heart rhl Na+ channel a subunits are expressed transiently in ts-21 cells. mino acid residues in transmembrane segment S6 of homologous domain IV (IVS6) of Na+ channel a subunits have important effects on block by local anesthetic drugs. lthough five amino acid residues in IVS6 differ between brain rii and cardiac rhl, exchange of these amino acid residues by sitedirected mutagenesis showed that only conversion of Thr-1755 in rhi to Val as in rii was sufficient to reduce the rate and extent of block by extracellular QX-314 and slow the escape of drug from closed channels after use-dependent bloclk Tetrodotoxin also reduced the rate of block by extracellular QX-314 and slowed escape of bound QX-314 via the extracellular pathway in rhl, indicating that QX-314 must move through the pore to escape. QX-314 binding was inhibited by mutation of Phe-1762 in the local anesthetic receptor site of rhl to la whether the drug was applied extracellularly or intracellularly. Thus, QX-314 binds to a single site in the rhl Na+ channel a subunit that contains Phe-1762, whether it is applied from the extracellular or intracellular side of the membrane. ccess to that site from the extracellular side of the pore is determined by the amino acid at position 1755 in the rhl cardiac Na+ channel. Cardiac Na+ channels carry the inward current during the initial rising phase of the cardiac action potential and are responsible for the rapid spread of excitation through the atria and ventricles that is required for synchronous cardiac contraction. The major component of voltage-sensitive Na+ channels from various tissues is a 26-kDa a subunit (reviewed in refs. 1 and 2). The primary sequence contains four homologous domains (I-IV), each containing six predicted transmembrane a-helices (S1-S6). The major Na+ channel subtype in the heart is distinguished functionally from other Na+ channels by its different kinetics and pharmacology. xpression of the a subunit of cardiac Na+ channels (rhl) (3, 4) in Xenopus oocytes (5, 6) or in mammalian cells (7, 8) produces Na+ channels with functional properties characteristic of cardiac cells. Voltage-sensitive Na+ channels are blocked by local anesthetic and antiarrhythmic drugs such as lidocaine in a voltageand/or frequency-dependent fashion (for review see ref. 9). lock is increased in depolarized and rapidly firing tissues. Drugs containing a tertiary amino group are present in neutral The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C solely to indicate this fact and charged forms at physiological ph, and the uncharged form is thought to equilibrate rapidly across cell membranes allowing these drugs to block Na+ channels when applied to either the inside or outside of the membrane (1). ecause permanently charged quaternary lidocaine derivatives such as QX-314 block neuronal Na+ channels from the intracellular but not from the extracellular side, intracellular block by tertiary compounds has been interpreted to be due to tertiary drugs passing through the membrane in their uncharged form to reach an intracellular binding site (11, 12). Recently, QX-314 was found to block native cardiac Na+ channels from the extracellular side (13), suggesting that cardiac channels differ from neuronal Na+ channels in the molecular mechanism by which they are blocked by local anesthetic and antiarrhythmic drugs. mino acids in predicted transmembrane segment 6 of homologous domain IV (IVS6) of the rat brain rii Na+ channel play critical roles in the action of local anesthetic drugs (14). Mutation of Phe-1764 near the center of this segment to la (F1764) virtually abolished use-dependent block by disrupting the binding of local anesthetics to depolarized channels. Mutation of Ile-176 near the extracellular end of IVS6 to la created a pathway by which extracellular QX-314, which is normally ineffective, could block the channel. The present study was designed to examine block of the cardiac Na+ channel by extracellular local anesthetic and antiarrhythmic drugs using QX-314 as a probe and to determine whether amino acid residues in transmembrane segment IVS6 are responsible for the difference in access of extracellular drugs to their receptor site in cardiac and brain Na+ channels. XPRIMNTL PROCDURS Construction and Mutagenesis of rhl Na+ Channel a-subunit cdn. full-length cdn encoding the rhl Na+ channel a subunit was assembled in luescript SK+ by PCR amplification using primers designed from the rhl sequence (refs. 3 and 7; J.R., T.T., and W..C., unpublished). The plasmid psp64t-rhl was created by insertion of the full-length rhl a-subunit cdn into psp64t (15). For transient expression in mammalian cells, the rhl a-subunit cdn was cloned into pcdm8 (Invitrogen), creating pcdm8-rhl. The plasmid pcdm8-rii was created by insertion of the rii sodium channel a-subunit cdn (16) into pcmd8. mutagenesis template for the rhl sodium channel was made by subcloning a stii/not I fragment (nucleotides ) of the rhl cdn into the corresponding sites of a modified phage M13mp19 (GICO/RL/Life Technologies) created by addition of stii and Not I recognition sites to the polylinker. Single-stranded template DN was prepared and oligonucleotide-directed mutagenesis was performed using standard protocols (17, 18). Mutant stii/not I fragments were then subcloned into pcdm8-rhl and/or psp64t-rh1 and confirmed by sequencing and qualitative restriction mapping. bbreviations: WT, wild-type; TTX, tetrodotoxin.

2 1184 Pharmacology: Qu et al. xpression of Na+ Channels and lectrophysiological Recordings from Xenopus Oocytes. The plasmids psp64t-rhi and pcdm8-rii were linearized with Not I and Cla I, respectively, and capped mrn was prepared using an in vitro transcription kit (mmessage mmachine; mbion, ustin, TX). The concentrations of mrn were calculated from percent incorporation of [a-32p]ctp and the quality of transcript was assessed by denaturing agarose gel electrophoresis. mrn was injected into Xenopus oocytes and Na+ currents were recorded as described (19). Mammalian Cell Transfection. Human ts-21 cells, an embryonic kidney cell line stably transfected with simian virus 4 large tumor antigen (Robert Dubridge, Cell Genesis, Foster City, C), were maintained in Dulbecco's modified agle medium/f12 medium (GICO/RL/Life Technologies) supplemented with 1% fetal bovine serum (HyClone), 25 units of penicillin per ml, and 25 jig of streptomycin per ml (Sigm. Cells were cotransfected with either wild-type (WT) or mutant sodium channel constructs and a vector encoding the human CD8 cell surface protein (O-pCD-leu2; merican Type Culture Collection) as described (2). Transiently transfected cells were visualized with magnetic polystyrene microspheres precoated with anti-cd8 antibody (Dynabeads M-45 CD8, Dynal, Great Neck, NY) (21). lectrophysiological Recording from Mammalian Cells. Whole-cell voltage-clamp experiments were performed on transiently transfected ts-21 cells at 22 C as described (7). QX-314 (N-ethyllidocaine bromide; Research iochemicals, Natick, M) was dissolved in either extracellular or intracellular solution at the final concentration and applied with a gravity-fed perfusion system (22). Solution changes were complete within 1-2 s. The intracellular (pipette) solution consisted of 9 mm CsF, 5 mm CsCl, 1 mm CsGT, 1 mm NaF, 2 mm MgCl2, and 1 mm Hepes, ph 7.4. The extracellular (bath) solution consisted of 14 mm NaCl, 5 mm CsCl, 1.8 mm CaC12, 1. mm MgCI2, 1 mm glucose, and 1 mm C: r.8 N.6 Z C: C-) 2 1. N.8.6 Z FIG. 1. lock of cardiac but not brain Na+ channels by extracellular QX-314 in ts-21 cells. ( and ) Development of block with time after exposure to.5 mm extracellular QX-314 (arrow) in cells expressing either brain () or cardiac () Na+ channels. Peak currents evoked by 16-ms-long pulses to -1 mv from a holding potential of -12 mv at.5 Hz were normalized to peak current in control and plotted versus elapsed time during the experiment. (Insets) Current traces recorded in control (larger) and 3 min after beginning QX-314 perfusion (smaller). Scale bars = 2 ms, 3 n. (C and D) Currentvoltage relationships recorded from the cells expressing brain (C) and cardiac (D) Na+ channels in control () and 5 min after beginning QX-314 perfusion (-). Proc. Natl. cad. Sci. US 92 (1995) Hepes, ph 7.4. Pooled data are reported as mean + SM. Statistical comparisons were made using Student's t test, and P <.5 was considered significant. RSULTS lock of Cardiac but Not rain Na+ Channels by xtracellular QX-314. The charged, membrane-impermeant quaternary lidocaine analog QX-314 has been a valuable tool for probing mechanisms of action of local anesthetic and antiarrhythmic drugs. Consistent with previous results in nerve cells (11, 12), exposure of cells expressing brain Na+ channels (rii; ref. 23) to.5 mm QX-314 resulted in no detectable decrease in current (Fig. 1 and C; mean = 6% + 6%, n = 7). xposure of four cells to concentrations up to 2 mm also produced no effect. In contrast, exposure of cells expressing cardiac Na+ channels (rh1; ref. 3) to.5 mm QX-314 resulted in 45% + 8% (n = 5) block of the Na+ current after 3 min of perfusion (Fig. 1 and D). For three cells studied after >1 min of perfusion, block was 82% ±+ 5%, corresponding to an C5 of.11 mm. lock of cardiac Na+ current by extracellularly applied QX-314 is consistent with previous results in native cardiac Purkinje fibers (13) and ventricular muscle cells (24). lock of rain and Cardiac Na+ Channels by Intracellular QX-314. fter achieving the whole-cell configuration, 1 min were allowed for QX-314 included in the recording pipette (2,uM) to diffuse into the cell. Then, a train of depolarizations (15 ms long, 1 Hz to -1 mv) was applied. Strong use-dependent block of brain and cardiac Na+ channels was C i 5~~in 1 a 1. (1) o ').8 - X O r 2, D < -D.5 N. Z FIG. 2. lock of brain and cardiac Na+ channels expressed in ts-21 cells by intracellular QX-314. train of 3, 15-ms-long pulses to -1 mv was applied at 1 Hz from a holding potential of -14 mv to produce use-dependent block. Then, identical depolarizations were applied once every 5 min to monitor the recovery of the current from drug block. ( and ) Normalized current traces recorded from brain () and cardiac () Na+ channels. The numbers indicate the 1st (1) and 3th (2) pulses of the 1-Hz train and currents recorded during pulses 5 min (3), 1 min (4), and 15 min (5) after the end of the train. (C) Time course of use-dependent block. Peak currents during the 1-Hz train were normalized to the peak current elicited by the first pulse of the train and plotted versus time. (D) Time course of recovery from block by intracellular QX-314. Currents were normalized by the difference between the peak current elicited by the 1st pulse and 3th pulses (I -I3oth)/(Ilst - I3th). Data in C and D are mean values + SM from three cells. The lines in D are exponentials with time constants obtained from similar experiments in which test pulses were applied more frequently.

3 observed (Fig. 2 and, traces 1 and 2). pulse-by-pulse decrease in current was observed in cells expressing either channel (Fig. 2C). In the absence of drug, such trains caused only.2% ± 2.6% (n = 3) reduction in peak current through cardiac channels (not shown). Thus, QX-314 can block brain and cardiac Na+ channels via an intracellular pathway. rain Na+ channels recover from block by QX-314 slowly with full recovery requiring many minutes (14), probably reflecting trapping of the charged drug in the closed channel (9, 1, 12, 25). To compare recovery from use-dependent block in brain and cardiac channels, a population of QX-314-blocked channels was produced using a train of depolarizations to -1 mv, and recovery at the holding potential of -14 mv was followed using infrequent short test pulses (Fig. 2 and, traces 3-5). Only 9% of the current in cells expressing the brain channels recovered during the 1-min monitoring period (Fig. 2D). In contrast, 89% of the current recovered within 5 min and recovery was virtually complete after 1 min at this holding potential for cells expressing the cardiac channels (Fig. 2D). Thus, recovery after use-dependent block by QX-314 is far more rapid for the cardiac than for the brain channel. mino cids Near the xtracellular nd of IVS6 Control an ccess Pathway for xtracellular QX-314. In mapping residues associated with the local anesthetic binding site, Ragsdale et al. (14) showed that mutation of Ile-176 to la near the extracellular end of transmembrane helix IVS6 created an access pathway that allowed extracellularly applied QX-314 to block the channel and speeded recovery after intracellular block by QX-314. The similarity of this mutant brain channel to the native cardiac channel suggested that the differences in access of extracellular QX-314 between cardiac and brain channels might be-explained by differences in channel structure near the extracellular end of IVS6. The amino acid sequences for the brain and cardiac Na+ channel a subunits differ from each other at five positions in transmembrane segment IVS6 (Fig. 3). To identify amino acids that were responsible for the different characteristics of drug block, we constructed mutant rhl channels in which these five nonconserved amino acids were changed to their 1. L-.8 L3 O.6 N =.4 o.2 z. Pharmacology: Qu et al * * 1774 cardiacvgilfftty1iisfllvvnmy I II L brain ---F--VS V V C v rhi, WT a RII * Ll 752F Tl 756S FIG. 3. ffects of mutations in Na+ channel segment IVS6 on recovery from block by QX-314 in Xenopus oocytes. () mino acid sequences of segment IVS6 of the rhl and RII Na+ channel a subunits. ( and C) Xenopus oocytes expressing the indicated Na+ channel constructs were microinjected with 5 nl of a 2 mm solution of QX-314 at least 1 min before recording. To develop use-dependent block, a 1-Hz train of twenty 15-ms-long pulses to -1 mv from a holding potential of -1 mv was applied to oocytes that had been preinjected with QX-314. Recovery from use-dependent block was monitored with test depolarizations applied once every min following the end of the train. Currents were normalized as described in Fig. 2D. ach data set contains mean values ± SM from three to seven oocytes. Proc. Natl. cad. Sci. US 92 (1995) brain counterparts. Consistent with the results in ts-21 cells, expression of the WT heart and brain Na+ channels inxenopus oocytes gave rapidly and slowly recovering currents, respectively, after use-dependent block by intracellularly applied QX-314 (Fig. 3). The single mutations 11764V and 11772V did not significantly affect recovery of the cardiac channel from drug block (Fig. 3). In contrast, mutation of three amino acids clustered near the extracellular end of IVS6 to their brain counterparts (LTT1752,1755,1756FVS) caused slowed recovery from use-dependent block, a phenotype more similar to that of the brain channel. To determine the contribution of each amino acid in this triplet, cdns in which each of the amino acids was mutated individually were constructed. xpression inxenopus oocytes showed that one mutant, T1755V, slowed recovery from block almost as much as LTT1752, 1755,1756FVS (Fig. 3C). In contrast, mutant L1752F had no detectable effect and T1756S had only a small effect on recovery (Fig. 3C). These results in Xenopus oocytes were confirmed when mutants LTT1752,1755,1756FVS and T1755V were studied after transient expression in ts-21 cells (Fig. 4). For both mutants, recovery from use-dependent block by QX-314 was substantially slowed in comparison to recovery in the wild-type cardiac Na+ channel. If the slower recovery from use-dependent block in mutant channels LTT1752,1755,1756FVS and T1755V was due to occlusion of an extracellular pathway for escape of QX-314 from the channel in these mutants, block by extracellular QX-314 ought to be impeded as well. Since the drug could dissociate from the mutant channels at a measurable rate, block by extracellular QX-314 was expected to be slowed but not completely prevented. To test this possibility,.5 mm QX-314 was applied to the outside of ts-21 cells expressing WT and mutant channels using a rapid perfusion system so that the rate of onset of extracellular block could be determined (Fig. 4). Native rhl channels were blocked rapidly. In contrast, block of mutant channels LTT1752,1755,1756FVS and T1755V was substantially slowed. The simplest interpretation of these results is that mutation T1755V impedes passage of QX-314 between the extracellular medium and its binding site, whether the drug is initially presented to the channel from the intracellular side (Fig. 4) or extracellular side (Fig. 4). xtracellular QX-314 pproaches Its Receptor Site Through the Pore. Tetrodotoxin (TTX) is thought to block the extracellular mouth of the pore of the Na+ channel (reviewed in refs. 1 and 2) by interacting with acidic residues in the short N I or,% DRII7- *LTT/FVS.5 * T1 755V I FiG. 4. ffects of mutations in transmembrane segment IVS6 expressed in ts-21 cells on block by QX-314. () Recovery from use-dependent block in ts-21 cells expressing rhl (), RII (L), WT channels, and mutants LTT1752,1755,1756FVS (-) and T1755V (-) of the rhl channel. The protocol was the same as that used in Fig. 2D (n = 3). () ffect of the same mutations on development of block by extracellularly applied QX-314. The protocol was the same as that used in Fig. 1. The points with error bars are mean values + SM from four to seven cells analyzed at 3 min.

4 11842 Pharmacology: Qu et al. 1. OD.8 N,x.6 Z FIG. 5. ffects of TTX on the rate of binding and release of extracellular QX-314. () Onset of block by 5,LM QX-314 in the absence () and presence (-) of 2,uM TTX. The protocol was the same as that used in Fig. 1. Due to use-dependent TTX block (29), TTX applied alone blocked -75% of the current at this pulse rate. ach point is mean ± SM from five or six cells. () Recovery from use-dependent block by 2,uM intracellular QX-314 in the absence (-) and presence (-) of 2,uM extracellular TTX. xtracellular TTX applied alone causes 35% additional use-dependent block of the current that recovers completely within 1 min (). The protocol was the same as that used in Fig. 2 except that recovery was monitored once every 3 s in the presence of TTX alone. TTX alone blocked 54% of the current at steady state with one pulse per min. segments SS2 between the S5 and S6 segments in each domain (refs ; for review see ref. 29). To test whether the extracellular access pathway in cardiac Na+ channels required QX-314 to move through the extracellular mouth of the pore, the rates of onset and reversal of QX-314 block were measured in the presence of TTX (2,tM), which blocks -5% of resting cardiac Na+ channels in a rapidly reversible manner (29). If QX-314 must pass through the outer mouth of the pore, TTX should reduce the rate of onset and reversal of QX-314 block by reducing the average fraction of Na+ channels whose drug receptor site is accessible to the extracellular medium. Consistent with this prediction, TTX reduces the rate of QX-314 block of Na+ current (Fig. S) and the rate of reversal after use-dependent block (Fig. 5). lthough consistent with our prediction, the amount of slowing of reversal is greater than predicted by a simple blocking model and suggests that these drugs may also interact to stabilize each other in their binding sites. Mutation of a Single mino cid Prevents lock by Intracellular and xtracellular QX-314. Mutant F1764 of the ril brain Na+ channel a subunit dramatically reduces the sensitivity of the channel to block by local anesthetics by reducing the affinity with which they bind to the channel (14). The equivalent mutation of the rhl a subunit, F1762, was.8 - ) 'L.6 z : * rh WT -1. n RII.9 - F1762 v LTT/FVS.8 - II* T1755V *rhl, W F1762 FI.5 -, FIG. 6. ffects of mutant F1762 of the rhl Na+ channel a subunit on block by QX-314. () Use-dependent block by intracellular QX-314 in the indicated mutants (n = 3). The protocol was the same as that used in Fig. 2C. () Time course of block during exposure to extracellular QX-314. The protocol was the same as that used in Fig. 1. Proc. Natl. cad. Sci. US 92 (1995) constructed and expressed in ts-21 cells. With intracellular application of.2 mm QX-314, cells expressing rhl, brain, LTT1752,1755,1756FVS, and T1755V were subject to extensive use-dependent block in response to a 1-Hz train of depolarizations to -1 mv (Fig. 6). In contrast, no usedependent block was observed for mutant F1762, and 99% of the current remained after 3 pulses (Fig. 6). lock by extracellularly applied QX-314 was also absent in this mutant (Fig. 6). The finding that this single mutation inhibits block by intracellularly and extracellularly applied QX-314 suggests that drugs applied by both pathways are interacting with the same site within the transmembrane pore of the channel. DISCUSSION Single Receptor Site for Intracellular and xtracellular QX-314. Our results support a model with a single binding site for QX-314 in the transmembrane pore of the Na+ channel (9, 1, 12), located on the intracellular side of Thr-1755 which controls drug movement between the binding site and the extracellular solution. The drug can reach this site readily from inside the cell when the channel is opened by trains of depolarizations that produce use-dependent block in rii and rhl Na+ channels. It can also reach this site slowly from the extracellular side of the rhl without channel opening. The finding that mutation of a single critical amino acid, Phe-1762, prevents block by intracellularly and extracellularly applied QX-314 confirms that intracellular and extracellular QX-314 interact with this single binding determinant. The simplest conclusion is that QX-314, whether applied extracellularly or intracellularly, binds at a single site containing Phe Our data are consistent with results showing that a component of block of batrachotoxin-modified cardiac Na+ channels in lipid bilayers by extracellularly and intracellularly applied lidocaine occurs at a single site but the drug arrives at that site by different access pathways (3). The cardiac action potential upstroke (31) and cardiac Na+ channels with intact inactivation (13) can be blocked by extracellular QX-314. Since extracellular, but not intracellular, application of QX-314 produced considerable tonic block of cardiac Na+ currents, two different blocking sites have been suggested (13). lternatively, these data can be explained by our proposal of a single binding site that is slowly accessible to QX-314 from the extracellular surface allowing tonic block, but that is not accessible to QX-314 from the intracellular surface until the channel is open. lock by extracellular QX-314 has not been observed after enzymatic removal of inactivation (32) or pharmacological disruption of inactivation with batrachotoxin (3). This is consistent with our single-site model since the rate of block by extracellular QX-314 is extremely slow (Fig. 1), whereas reversal of block in the absence of inactivation is predicted to be rapid once the channel is opened due to escape of drug to the intracellular compartment. vidence for an additional drug-receptor complex was also observed in our experiments. t unphysiologically high stimulation frequencies (e.g., 2 Hz), use-dependent block by extracellularly applied QX-314 was observed in rhl and ril Na+ channels. Recovery from this extracellular use-dependent block was extremely rapid (<2 s for complete recovery), distinguishing it from use-dependent block by intracellular QX-314. Further experiments will be necessary to characterize the binding site responsible for this additional block and its relationship to block at more physiological frequencies of stimulation. Thr-1755 Controls Passage of xtracellular Drug Through the Pore to the ntiarrhythmic Drug Receptor Site. The pathway for reaching the drug receptor site from the extracellular side is available in the rhl cardiac channel but not in the rii brain channel. In the rii channel, mutation 1176

5 Pharmacology: Qu et al. created a phenotype similar to that of the cardiac channel (14). lthough this residue is conserved in the rhl channel (11758), QX-314 can still block from the extracellular side. Instead, Thr-1755 seems to create the extracellular access pathway in rhl since mutation T1755V produces a brain-like phenotype in the cardiac channel. The effect of the mutation in the rhl channel is unknown since expression of this construct did not produce functional channels. Residues at positions 1755 and 1758 in the rhl a subunit primary sequence are expected to fall on the same face of the predicted IVS6 a-helix, and the residues at position 1755 in rhl and at the position analogous to 1758 in rii (position 176) both influence QX-314 access to its binding site from the extracellular side. Residue F1762 controlling QX-314 affinity is also predicted to lie on the same surface of the a-helix. ecause QX-314 has been shown to bind in the Na+ channel pore (33,34), the simplest interpretation is that residues 1755 and 1758 in rhl face the pore and affect bidirectional access of extracellular QX-314 to and from the binding site containing F1762. Consistent with that interpretation, TTX slows dissociation of QX-314 through the extracellular access pathway. Since TTX is well characterized as an extracellularly acting pore blocker (for review see ref. 29), this result provides strong support for the idea that QX-314 moves through the pore in reaching its receptor site. Of the amino acids in the rhl channel that differed from the ril isoform, only T1755V affected block by QX-314. Since this mutation was not sufficient by itself to reproduce the RII phenotype completely, additional differences in channel structure must be responsible for the remaining differences in phenotype. These other regions are likely to include S6 segments of homologous domains I-III as well as the predicted pore region of each homologous domain located between the S5 and S6 transmembrane segments. Further mutations in these regions will be required to account for the remaining differences in QX-314 block between the brain and cardiac isoforms. We thank Dr. ertil Hille for critical comments on a draft of the manuscript. This work was supported by National Institutes of Health Research Grant P1-HL44948 to W..C. and a postdoctoral research fellowship from the merican Heart ssociation (Washington ffiliate) to Y.Q. 1. Catterall, W.. (1992) Physiol. Rev. 72, S15-S Cohen, S.. & archi, R. L. (1993) Int. Rev. Cytol. 137C, Rogart, R.., Cribbs, L. L., Muglia, L. K., Kephart, D. D. & Kaiser, M. W. (1989) Proc. Natl. cad. Sci. US 86, Kallen, R. G., Sheng, Z.-H., Yang, J., Chen, L., Rogart, R.. & archi, R. L. (199) Neuron 4, Cribbs, L. L., Satin, J., Fozzard, H.. & Rogart, R.. (199) FS Lett. 275, Proc. Natl. cad. Sci. US 92 (1995) White, M. M., Chen, L. Q., Kleinfield, R., Kallen, R. G. & archi, R. L. (1991) Mol. Pharmacol. 39, Qu, Y., Rogers, J., Tanada, T., Scheuer, T. & Catterall, W.. (1994) Proc. Natl. cad. Sci. US 91, Chahine, M., ennett, P.., George,. L., Jr., & Horn, R. (1994) Pflugers rch. 427, utterworth, J. F. & Strichartz, G. R. (199) nesthesiology 72, Hille,. (1977) J. Gen. Physiol. 69, Frazier, D. T., Narahashi, T. & Yamada, M. (197) J. Pharmacol. xp. Ther. 171, Strichartz, G. R. (1973) J. Gen. Physiol. 62, lpert, L.., Fozzard, H.., Hanck, D.. & Makielski, J. C. (1989) m. J. Physiol. 257, H79-H Ragsdale, D. S., McPhee, J. C., Scheuer, T. & Catterall, W.. (1994) Science 265, Krieg, P.. & Melton, D.. (1984) Nucleic cids Res. 12, uld, V. J., Goldin,. L., Krafte, D. S., Catterall, W.., Lester, H.., Davidson, N. & Dunn, R. J. (199) Proc. Natl. cad. Sci. US 87, Kunkel, T.. (1985) Proc. Natl. cad. Sci. US 82, usubel, F. M., rent, R., Kingston, R.., Moore, D.., Seidman, J. G., Smith, J.. & Struhl, K. eds. (1994) Short Protocols in Molecular iology (Wiley Interscience, New York). 19. McPhee, J. C., Ragsdale, D. S., Scheuer, T. & Catterall, W.. (1994) Proc. Natl. cad. Sci. US 91, Margolskee, R. F., McHendry-Rinde,. & Horn, R. (1993) iotechniques 15, Jurman, M.., oland, L. M., Liu, Y. & Yellen, G. (1994) iotechniques 17, Friel, D. D. & ean,. P. (199) Pflugers rch. 415, uld, V. J., Goldin,. L., Krafte, D. S., Marshall, J., Dunn, J. M., Catterall, W.., Lester, H.., Davidson, N. & Dunn, R. J. (1988) Neuron 1, aumgarten, C. M., Makielski, J. C. & Fozzard, H.. (1991) J. Mol. Cell. Cardiol. 23, Suppl. I, Starmer, C. F., Yeh, J. Z. & Tanguy, J. (1986) iophys. J. 49, Terlau, H., Heinemann, S. H., Stuhmer, W., Pusch, M., Conti, F., Imoto, K. & Numa, S. (1991) FS Lett. 293, Satin, J., Kyle, J. W., Chen, M., ell, P., Cribbs, L. L., Fozzard, H.. & Rogart, R.. (1992) Science 256, ackx, P. H., Yue, D. T., Lawrence, J. H., Marban,. & Tomaselli, G. F. (1992) Science 257, Lipkind, G. M. & Fozzard, H.. (1994) iophys. J. 66, Zamponi, G. W., Doyle, D. D. & French, R. J. (1993) iophys. J. 65, Gintant, G.., Hoffman,. F. & Naylor, R.. (1983) Circ. Res. 52, Gingrich, K. J., eardsley, D. & Yue, D. T. (1993) J. Physiol. (London) 471, Cahalan, M. D. (1978) iophys. J. 23, Yeh, J. Z. & Oxford, G. S. (1985) J. Gen. Physiol. 85,