Modulation of Cardiac Na+ Channel Expression in Xenopus Oocytes by β1 Subunits

Voltage-gated Na+ channels consist of a large α subunit of 260 kDa associated with β1 and/or β2 subunits of 36 and 33 kDa, respectively. α subunits of rat cardiac Na+ channels (rH1) are functional when expressed alone in Xenopus oocytes or mammalian cells. β1 subunits are present in the heart, and localization of β1 subunit mRNA by in situ hybridization shows expression in the perinuclear cytoplasm of cardiac myocytes. Coexpression of β1 subunits with rH1 α subunits in Xenopus oocytes increases Na+ currents up to 6-fold in a concentration-dependent manner. However, no effects of β1 subunit coexpression on the kinetics or voltage dependence of the rH1 Na+ current were detected. Increased expression of Na+ currents is not observed when an equivalent mRNA encoding a nonfunctional mutant β1 subunit is coexpressed. Our results show that β1 subunits are expressed in cardiac muscle cells and that they interact with α subunits to increase the expression of cardiac Na+ channels in Xenopus oocytes, suggesting that β1 subunits are important determinants of the level of excitability of cardiac myocytes in vivo.

Voltage-gated Na ؉ channels consist of a large ␣ subunit of 260 kDa associated with ␤1 and/or ␤2 subunits of 36 and 33 kDa, respectively. ␣ subunits of rat cardiac Na ؉ channels (rH1) are functional when expressed alone in Xenopus oocytes or mammalian cells. ␤1 subunits are present in the heart, and localization of ␤1 subunit mRNA by in situ hybridization shows expression in the perinuclear cytoplasm of cardiac myocytes. Coexpression of ␤1 subunits with rH1 ␣ subunits in Xenopus oocytes increases Na ؉ currents up to 6-fold in a concentration-dependent manner. However, no effects of ␤1 subunit coexpression on the kinetics or voltage dependence of the rH1 Na ؉ current were detected. Increased expression of Na ؉ currents is not observed when an equivalent mRNA encoding a nonfunctional mutant ␤1 subunit is coexpressed. Our results show that ␤1 subunits are expressed in cardiac muscle cells and that they interact with ␣ subunits to increase the expression of cardiac Na ؉ channels in Xenopus oocytes, suggesting that ␤1 subunits are important determinants of the level of excitability of cardiac myocytes in vivo.
Cardiac Na ϩ channels are responsible for the rapid, depolarizing upstroke in the cardiac action potential. Their function is critical for the rapid spread of depolarization through the heart and, ultimately, for cardiac contractility. Cardiac Na ϩ channels have properties that distinguish them from other well characterized voltage-dependent Na ϩ channels. They are less sensitive to tetrodotoxin (Baer et al., 1976), and their kinetics of activation and inactivation are slower and more complex (Brown et al., 1981). A Na ϩ channel ␣ subunit cDNA encoding this channel has been isolated from newborn rat heart (rH1) 1 (Rogart et al., 1989) and denervated rat skeletal muscle (Skm2, Kallen et al. (1990)) and a closely related Na ϩ channel has been isolated from human cardiac muscle (Gellens et al., 1992). Expression of these ␣ subunit cDNAs in Xenopus oocytes (Cribbs et al., 1990;White et al., 1991;Gellens et al., 1992) and mammalian cells (Qu et al., 1994;O'Leary and Horn, 1994) yields Na ϩ currents with functional properties and tetradotoxin sensitivity characteristic of the native cardiac Na ϩ channel.
In electric eel electroplax, Na ϩ channels are composed of a single large ␣ subunit with a molecular mass of 230 -270 kDa (Agnew et al., 1980;Miller et al., 1983;Norman et al., 1983). The major form of the Na ϩ channel in rat brain is a heterotrimeric complex of an ␣ subunit (260 kDa), a noncovalently bound ␤1 subunit (36 kDa), and a disulfide-linked ␤2 subunit (33 kDa) (Catterall, 1992). Na ϩ channels in rat skeletal muscle are heterodimeric, composed of an ␣ subunit and only one ␤ subunit (Barchi, 1983;Kraner et al., 1985) which is encoded by the same gene as the brain ␤1 subunit (Makita et al., 1994). Currents due to brain or skeletal muscle Na ϩ channel ␣ subunits expressed alone by injection of mRNA in Xenopus oocytes are small and have abnormally slow kinetics (Auld et al., 1988;Krafte et al., 1988;Joho et al., 1990;Krafte et al., 1990). Coexpression of the ␤1 subunit increases channel expression, shifts the gating mode from slow to fast, speeds activation and inactivation kinetics, and causes a hyperpolarizing shift in the voltage dependence of activation and inactivation (Isom et al., 1992;Cannon et al., 1993;Makita et al., 1994;Patton et al., 1994;Isom et al., 1995). Thus, ␤1 subunits both modify the functional properties of brain and skeletal muscle Na ϩ channels and increase the efficiency of their expression.
The role of the ␤1 subunit in the heart has been less clear. Initial studies using subunit-specific antibodies identified the ␤1 subunit polypeptide in the heart (Sutkowski and Catterall, 1990). ␤1 subunit mRNA also has been identified in heart (Isom et al., 1992;Tong et al., 1993;Yang et al., 1993;Makita et al., 1994) and cDNAs homologous to the rat brain ␤1 transcript have been cloned from rat cardiac muscle McClatchey et al., 1993;Yang et al., 1993). The human and rat cardiac ␤1 cDNAs are identical in sequence to their brain counterparts (Makita et al., 1994). 2 However, purified preparations of cardiac Na ϩ channels from chicken and rat do not have associated ␤1 subunits (Lombet and Lazdunski, 1984;Cohen and Levitt, 1993), and there have been conflicting reports concerning the functional significance of ␤1 subunits in the heart. Yang et al. (1993) reported no effect of ␤1 expression, while Kyle et al. (1993) observed a depolarizing shift in the voltage dependence of inactivation as a result of ␤1 coexpression.
In these experiments, we have investigated whether ␤1 subunit mRNA is expressed in cardiac muscle cells using in situ hybridization, and we have examined the functional role of ␤1 subunits in modulating cardiac Na ϩ channel expression and kinetics by coexpression of mRNA for ␣ and ␤1 subunits in Xenopus oocytes. We report that ␤1 subunit mRNA is expressed in cardiac muscle cells in vivo. When expressed in Xenopus oocytes in conjunction with the rH1 ␣ subunit, ␤1 subunits substantially increase Na ϩ channel expression, but have no detectable effects on physiological properties. Thus, ␤1 subunits are likely to be important determinants of the level of functional expression of Na ϩ channels in cardiac cells, but do not alter their physiological properties significantly.

EXPERIMENTAL PROCEDURES
Construction of Expression Vectors-A full-length cDNA encoding the rH1 Na ϩ channel ␣ subunit was assembled in pBluescript SK ϩ by polymerase chain reaction amplification and sequencing of cDNA fragments using primers designed from the rH1 sequence (Rogart et al., 1989;Qu et al., 1994). 3 The full-length ␣ subunit cDNA was then cloned into the BglII site of pSP64T. The oocyte expression vector p␤1.SP64T was constructed as described previously (Patton et al., 1994). Both vectors contained only the coding sequence of the cDNAs.
The deletion mutant ␤1 ⌬Val138-Ser159 was constructed by first removing a portion of the 5Ј-untranslated region of p␤1.C1Aa (Isom et al., 1992) by deletion of nucleotides 1 through 175. This region is predicted to contain stem-loop structures which may decrease the efficiency of ␤1 expression (Patton et al., 1994). Single stranded DNA was prepared by interference helper phage VCS-M13 infection of XL1-blue transformants (Stratagene, La Jolla, CA), and served as template for oligonucleotide-directed "loop out" deletion mutagenesis. A 36-base "clamp" oligonucleotide was designed to anneal to two segments of the sense strand template which flanked the region to be deleted. In vitro mutagenesis reactions were performed as described for the Sculptor TM IVM system (Amersham Corp.). The resulting mutant phagemid contained a 66-base pair deletion of the nucleotides encoding Val 138 through Ser 159 . The deleted amino acids are located in the extracellular domain of the ␤1 polypeptide, adjacent to the proposed transmembrane segment, and do not contain a putative glycosylation site.
mRNA Transcription-pSP64T.rH1.2 was linearized with NotI and p␤1.C1Aa was linearized with EcoRI and transcription was performed using the Ambion SP6 mMessage mMachine kit according to the manufacturer's instructions. The quality of the RNA from each transcription reaction was checked by agarose gel electrophoresis, and the RNA dissolved in nuclease-free water for oocyte injection.
Expression of Constructs and Electrophysiological Recording-Ovarian lobes were removed from adult female Xenopus laevis and the follicular layers were digested using collagenase (2.3 mg/ml, Sigma, Type I) dissolved in OR-2 (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl 2 , 5 mM HEPES, pH 7.5). After digestion, healthy oocytes were manually selected based upon size and uniformity of color and incubated at 18°C in Barth's medium (88 mM NaCl, 1 mM KCl, 0.82 mM MgSO 4 , 0.33 mM Ca(NO 3 ) 2 , 0.41 mM CaCl 2 , 2.4 mM NaHCO 3 , 10 mM HEPES, pH 7.4, 50 g/ml gentamicin, and 5% fetal bovine serum). Twenty-four h after collagenase digestion, oocytes were again selected and injected with the mRNA mentioned above. The oocytes were studied 3 days after injection.
For two-microelectrode voltage-clamp experiments, the oocytes were continuously perfused at room temperature (23-25°C) with Ringer's solution (115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl 2 , 10 mM HEPES, pH 7.2). Whole cell Na ϩ currents were studied using the conventional two-microelectrode voltage-clamp technique (Patton and Goldin, 1991) with a CA-1 amplifier (DAGAN). For cell-attached macropatch recordings, oocytes were first manually stripped with fine forceps under a dissecting microscope after shrinking with a hypertonic solution (200 mM potassium aspartate, 20 mM KCl, 10 mM EGTA, 1 mM MgCl 2 , 20 mM HEPES, pH 7.4 -7.5; Methfessel et al. (1986)). During recordings, the oocytes were bathed in a high K ϩ solution (110 mM KCl, 10 mM NaCl, 10 mM EGTA, 1 mM MgCl 2 , 10 mM HEPES, pH 7.2) to bring the membrane potential to approximately 0 mV. The electrode tip was coated with Sylgard (Dow Corning) and filled with 150 mM NaCl, 1.5 mM CaCl 2 , 2 mM MgCl 2 , 5 mM HEPES, pH 7.4 (tip resistance 0.5-1 m⍀). Macropatch current was recorded in the cell-attached configuration (Hamill et al., 1981) using an AxoPatch-1C amplifier (Axon Instruments). The voltage-clamp protocols are described in figure legends or corresponding text. Conductance-voltage (g-V) relationships were calculated from current-voltage (I-V) relationships according to g ϭ I/(V-V r ), where I is the peak current measured at voltage V and V r is the measured reversal potential. Normalized conductance-voltage relationships and inactivation curves were fit with a Boltzmann distribution, 1/(1 ϩ exp[(V-V1 ⁄2 )/ k]), where V1 ⁄2 is the voltage of half-activation or half-inactivation and k is a slope factor. Pooled data are reported as means Ϯ S.E. Statistical comparisons were made using Student's t test, with p Ͻ 0.05 taken as the critierion of significance.
In Situ Hybridization Analysis of Rat Cardiac ␤1 Subunit mRNA-A probe for in situ hybridization analysis of cardiac ␤1 subunit mRNA expression was generated by polymerase chain reaction using the following conditions: 1 ng of human genomic DNA template, 4 mM each primer (forward: GGGCTGCGTGGAGGTGG, reverse: TCTTGTGCAG-CAGCTTC), 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgSO 4 , 0.001% (w/v) gelatin, 200 mM each deoxyribonucleotide triphosphate, and 2.5 units of Taq DNA polymerase (Cetus). The amplification was performed as follows: 1 min at 90°C; 1 min 15 s at 50°C; 2 min at 72°C for 40 cycles followed by 10 min at 72°C and stopping of the reaction at 4°C. The resulting 527-nucleotide polymerase chain reaction product was sequenced using Sequenase (U. S. Biochemical Corp.), subcloned into pBluescript SK ϩ , and linearized with the appropriate restriction endonucleases to generate sense and antisense probe templates, respectively. This sequence was identical to that reported previously for rat brain ␤1 (Isom et al., 1992) except for the following changes: A for G at residue 312, T for A at residue 484, T for C at residue 629, and T for G at residue 753. Transcription reactions were carried out with either T3 or T7 RNA polymerase incorporating digoxigenin-11-UTP and quantitated according to the Genius System product literature (Boehringer Mannheim).
In situ hybridization of free-floating sections was carried out using modifications of methods described previously (Miller et al., 1989;Black et al., 1994). Briefly, adult Sprague-Dawley rats were anesthetized using sodium pentabarbitol, the heart was removed, immediately frozen in powdered dry ice, and stored at Ϫ70°C. Sagittal sections (40 m) through the long axis of the heart were cut on a sliding microtome and then placed into 4% paraformaldehyde fixative for 45 min. Tissue sections were then rinsed in 0.05 M phosphate-buffered saline for 20 min, water for 2 min, 0.1 M triethanolamine, pH 8.0, for 2 min and then in 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min to reduce nonspecific binding. The tissue sections were then rinsed in 2 ϫ SSC for 10 min, 70% ethanol for 2 min, 95% ethanol for 2 min, 100% ethanol for 2 min, 70% ethanol for 2 min, and finally in water for 2 min. Sections were prehybridized for 3 h at 42°C in a buffer containing 44.6% formamide, 8.9% dextran sulfate, 0.27 M NaCl, 7 mM Tris, pH 8.0, 0.7 mM EDTA, 0.9 ϫ Denhardt's, 16 mM dithiothreitol, 0.45 mg/ml yeast tRNA, and 4.6 ϫ Genius Northern blocking solution (Boehringer Mannheim). The tissue was then transferred to hybridization solution containing 41% formamide, 8.2% dextran sulfate, 0.25 M NaCl, 6.5 mM Tris, pH 8.0, 0.66 mM EDTA. 0.82 ϫ Denhardts, 14.8 mM dithiothreitol, 1.2 g/ml yeast tRNA, 4.1 ϫ Genius Northern blocking solution, and 7.5 g/ml of the digoxigenin-labeled probe and incubated overnight at 42°C. Tissue sections were rinsed in 1 ϫ standard sodium chloride/sodium citrate (SSC; Miller et al. 1989) for 30 min, treated with 25 g/ml RNase A in RNase buffer (10 mM Tris, 0.5 M NaCl, and 1 mM EDTA) for 30 min, rinsed in 1 ϫ SSC for 30 min, rinsed in 0.1 ϫ SSC at 45°C for 40 min, rinsed in 0.1 ϫ SSC at room temperature for 15 min, and then in 0.1 M Tris-buffered saline, pH 7.4, for 30 min. The tissue was then blocked using 10% normal sheep serum in 0.1 M Tris-buffered saline for 1 h at room temperature before being placed in alkaline phosphatase-conjugated anti-digoxigenin F(ab) antibody (diluted 1:400 in 0.1 M Trisbuffered saline containing 10% normal sheep serum and 0.1% Triton X-100) for 24 h at room temperature. The tissue was then washed in 0.1 M Tris-buffered saline for 1 h followed by 1 h of rinsing in a solution containing 100 mM Tris, 50 mM MgCl 2 , and 100 mM NaCl, pH 9.5. The tissue was then reacted with 4-nitro blue tetrazolium chloride (0.45 mg/ml), 5-bromo-4-chloro-3-indolyl phosphate (0.175 mg/ml), and levamisole (0.24 mg/ml) in the reaction buffer, rinsed in stop buffer (10 mM Tris and 1 mM EDTA, pH 8.0) for 30 min and the free floating sections were finally mounted on gelatin-subbed slides, air dried, coverslipped using Biomeda gel mount (Fischer), and viewed using a Leitz Dialux microscope.
Control experiments included omitting probes from hybridization solution and substituting sense probes for antisense probes. In addition, ␤1 antisense probe was hybridized to liver tissue to test for nonspecific labeling with the probe. No specific labeling was observed in the controls.

RESULTS
In Situ Hybridization of ␤1 in Heart-␤1 subunit mRNA and protein have been detected in the heart using immunoblotting and Northern blotting methods (Sutkowski and Catterall, 1990;Isom et al., 1992;Tong et al., 1993;Yang et al., 1993;Makita et al., 1994), but it is not known whether this ␤1 mRNA is in the neurons resident in the heart or in the cardiac myo-cytes themselves. The cellular distribution of sodium channel ␤1 subunit mRNA was examined in heart using non-isotopic in situ hybridization methods. At high magnification there is positive staining of muscle cells in regions of the cytoplasm surrounding the nucleus (Fig. 1, A-C). No staining was observed when the sense probe was used or when no probe was added to the hybridization solution (Fig. 1D). As a further control, liver tissue was also stained with the ␤1 probe and, as expected, there was no staining. Lower magnification views show that ␤1 is expressed in myocytes throughout the heart with the most prominent staining located in the papillary muscles of the ventricles. These results demonstrate the presence of ␤1 mRNA within most cardiac muscle cells.
Coinjection of Rat Brain ␤1 Subunit mRNA Increased Cardiac Na ϩ Channel Expression in Xenopus Oocytes-To investigate the functional consequences of coexpression of ␤1 subunits with cardiac rH1 ␣ subunits, we injected mRNAs encoding the two subunits into Xenopus oocytes. Na ϩ currents in the oocytes were measured 72 h after mRNA injection using two-microelectrode voltage clamp. At a single concentration of ␣ subunit, Na ϩ currents were dramatically larger at all test potentials in oocytes coinjected with ␤1 subunit mRNA (Fig. 2).
To examine the concentration dependence of this increase in Na ϩ current amplitude, ␤1 RNA was injected at concentrations ranging from 10 to 500 ng/l with a constant 25 ng/l of ␣ subunit RNA. The amplitude of the expressed Na ϩ current increased as the amount of ␤1 RNA injected was increased (Fig. 3A). No current was observed when ␤1 subunits were injected in the absence of the ␣ subunit. These results were obtained in one series of oocytes injected simultaneously and studied at times ranging from 62 to 78 h after injection. Similar results were obtained in two other experimental series of this kind.
Significant increases in Na ϩ channel current amplitude were observed at a ␤1:␣ RNA molar ratio of 5:3. The observed increase in current amplitude saturated at a ␤1:␣ RNA molar ratio of 50:1, with a half-maximal increase at a ␤1:␣ molar ratio of 10:1 (equivalent to a weight ratio of 2:1). The maximal increase in current induced by ␤1 subunit RNA coinjection was 3-6-fold, depending on the batch of oocytes (mean ϭ 5.2 Ϯ 1.4-fold, n ϭ 22 oocytes). This is comparable to the increase observed when rat brain type IIA Na ϩ channel ␣ subunits were coinjected with ␤1 subunit mRNA in Xenopus oocytes (Isom et al., 1992) and to the increase in [ 3 H]saxitoxin binding observed as a result of type IIA ␣ subunit and ␤1 subunit coexpression in transfected mammalian cells (Isom et al., 1994).
An alternative explanation for the observed increase in Na ϩ current amplitude is that ␤1 subunit mRNA may exert nonspecific effects on Na ϩ channel expression or mRNA expression in general. To examine that possibility, a deletion mutant of rat brain ␤1 (␤1⌬Val 138 -Ser 159 ) was constructed as described under "Experimental Procedures." The functional effects of this mutant ␤1 subunit were tested by coinjecting it with the rat brain type IIA ␣ subunit (␤1:␣ molar ratio of 100:1). The macroscopic Na ϩ current time courses from oocytes injected with transcripts encoding the rat brain type IIA ␣ subunit alone and in conjunction with ␤1⌬Val 138 -Ser 159 were very similar. Current-voltage relationships, steady-state inactivation curves, and the time course of recovery from inactivation also were unaffected (data not shown). This indicates that ␤1⌬Val 138 -Ser 159 is inactive, probably because it does not associate with the ␣ subunit. Coinjection of ␤1⌬Val 138 -Ser 159 with rH1 ␣ subunit mRNA (10:1 molar ratio) caused no increase in current amplitude (Fig. 3B). Wild-type ␤1 subunits caused the normal increase in current amplitude in the same batch of oocytes (Fig.  3B). Similar results were observed in two other experiments.
FIG. 1. In situ hybridization of ␤1 mRNA in heart. Adult rat heart sections were processed for in situ hybridization as described under "Experimental Procedures." A, ventricular papillary heart tissue hybridized with ␤1 antisense probe demonstrating the presence of ␤1 mRNA in muscle cells. Arrowheads outline the nuclei. B and C, higher magnifications of heart tissue hybridized with ␤1 antisense probe illustrating that labeling is present in the cytoplasm surrounding the nucleus. D, tissue section hybridized with ␤1 sense probe to illustrate the lack of hybridization and the specificity of the ␤1 antisense probe. Scale bars: A equals 20 m; B-D equal 10 m.
Thus, an active ␤1 subunit is necessary to observe the increase in current with ␤1 subunit RNA coinjection.
Coinjection of ␤1 Subunit RNA Did Not Change the Electrophysiological Properties of rH1 Na ϩ Current-Electrophysiological effects of the ␤1 subunit were examined by recording currents in Xenopus oocytes due to injection of the rH1 ␣ subunit RNA alone or in combination with ␤1 subunit mRNA. To insure adequate time and voltage control, recordings were made of macroscopic currents in small membrane patches (macropatches) that allow excellent time and spatial control of membrane potential. The effect of ␤1 subunits on expression of functional Na ϩ channels was verified by contemporaneous twomicroelectrode voltage clamp recordings from oocytes in the same batch used for kinetic comparisons. An average increase in current amplitude of 3-4-fold was observed in these ␤1coinjected oocytes. Fig. 4A shows macropatch recordings from oocytes injected with ␣ subunit RNA alone (500 ng, left panel) or in conjunction with ␤1 subunit RNA (100 ng of ␣, 200 ng of ␤1, right panel). Little overall difference in Na ϩ current kinetics was observed. To detect quantitative effects on current time course, normalized and averaged currents measured during test pulses to Ϫ60, Ϫ40, Ϫ20, or ϩ20 mV were compared (Fig.  4B). Average time courses obtained with (dotted line) and without (solid line) coinjection of ␤1 subunit mRNA were virtually superimposable. Inactivation time courses at ϩ20 mV with and without ␤1 were also compared by fitting the decaying phase of the current with 2 exponentials. In oocytes injected with ␣ subunit RNA alone, the time constant of the fast component (1) was 0.42 Ϯ 0.12 ms with an amplitude of 93% and the time constant of the slow component (2) was 3.4 Ϯ 1.6 ms (n ϭ 7).
FIG. 2. Effect of rat brain Na ؉ channel ␤1 subunit mRNA coinjection with rH1 ␣ subunit mRNA on expressed Na ؉ currents in Xenopus oocytes. Currents were recorded in oocytes injected with ␣ alone (25 ng/l, left) or ␣ (25 ng/l) ϩ ␤1 (50 ng/l, right) mRNA during pulses from a holding potential of Ϫ100 mV to potentials ranging from Ϫ45 to 0 mV in 5-mV increments using two-microelectrode voltage clamp.
FIG. 3. Effect of ␤1 subunit mRNA coinjection on rH1 Na ؉ current expression. Mean current levels were measured 72 h after injection in a single batch of oocytes (n ϭ 6 for each point). A, effects of ␤1 mRNA concentration. Each point was significantly different from the adjacent one (p Ͻ 0.05) except for the last two right-hand bars with the highest ␤1 mRNA concentrations. B, comparison of effects of wild type and mutant ␤1 subunit on Na ϩ channel expression. rH1 ␣ mRNAs was injected into oocytes with or without ␤1 mRNA or with an inactive mutant ␤1 mRNA (⌬Val 138 -Ser 159 ) with those amino acids deleted. The amounts of ␣ and ␤1 mRNAs injected are indicated beneath the histogram bars.
FIG. 4. Effects of coinjection of ␤1 subunit mRNA on the time course of rH1 Na ؉ current. A, current traces recorded in cell-attached macropatch configuration from oocytes injected with ␣ subunit mRNA alone (500 ng/l, left) and in oocytes coinjected with ␤1 subunit mRNA (␣ 100 ng/l ϩ ␤1 200 ng/l, right) during depolarizations to Ϫ80, Ϫ70, Ϫ60, Ϫ50, Ϫ40, Ϫ30, Ϫ20, Ϫ10, 0, 20, 60, and 80 mV from a holding potential of Ϫ120 mV. B, averages of macropatch current traces during pulses to the indicated voltages from oocytes injected with rH1 ␣ subunit mRNA alone (solid lines, n ϭ 18) or coinjected with ␤1 subunit mRNA (dotted lines, n ϭ 15). The average current traces shown were constructed by normalizing the amplitudes of current traces from individual experiments and then averaging them.
In oocytes coinjected with ␤1 subunit mRNA, 1 was 0.48 Ϯ 0.08 ms with an amplitude of 94% and 2 was 2.9 Ϯ 1.5 ms (n ϭ 8). An unpaired Student's t test showed no significant difference with all three parameters (p Ͼ 0.05). Despite having large effects on Na ϩ current amplitude, coinjection of ␤1 subunits had little effect on current time course.
Effects of ␤1 subunits on the voltage dependence of Na ϩ channel activation and inactivation were determined using both two-microelectrode and cell-attached patch recording configurations. Mean voltage dependence of Na ϩ current activation was unaffected by coexpression of the ␤1 subunit with the ␣ subunit in cell-attached patches or in two-microelectrode recordings (Fig. 5, A and B). The voltage dependence of Na ϩ channel inactivation was determined using conditioning prepulses (98 ms long in cell-attached experiments, 500 ms long in two-microelectrode experiments) followed by a test depolarization. The mean voltage dependences of Na ϩ channel activation and inactivation were not significantly affected in either recording configuration due to coinjection of ␤1 subunit RNA (Fig. 5, A and B). The voltage dependences of activation and inactivation determined in the cell-attached configuration with or without ␤1 subunits were shifted negatively compared to two-microelectrode recordings as has been previously observed for the native cardiac channel (Kimitsuki et al., 1990).
Recovery from inactivation was also studied in both the cell-attached and two-microelectrode recording configurations using double-pulse protocols. In the cell-attached configuration, Na ϩ channels were inactivated using a 16-ms conditioning pulse to Ϫ30 mV. The membrane potential was then returned to Ϫ120 mV to allow channels to recover from inactivation. The degree of recovery was assessed at various recovery times with a test pulse to Ϫ30 mV. Recovery time courses at Ϫ120 mV determined by plotting normalized peak test pulse current versus recovery time were well fit with single exponentials (Fig. 5C). In oocytes injected with ␣ subunit RNA alone, ϭ 6.9 Ϯ 1.8 ms, and in oocytes coinjected with ␤1 subunit mRNA, ϭ 7.3 Ϯ 1.4 ms (n ϭ 4). For recovery experiments in the two-microelectrode configuration, the conditioning pulse was 5 s long and recovery was studied at Ϫ100 mV. Such a long prepulse is expected to generate both fast and slow inactivation of the Na ϩ channel. After such prepulses, fits of the recovery time course required two exponential components (Fig. 5D). In oocytes injected with ␣ alone, the time constant of the fast component (1) was 5.5 Ϯ 3.1 ms, and the time constant of the slow component (2) was 1010 Ϯ 174 ms, with initial amplitudes, A1 and A2, of 52 and 48% respectively (n ϭ 3). In oocytes coinjected with ␤1 subunit mRNA, 1 was 6.0 Ϯ 0.9, 2 was 868 Ϯ 163, with A1 and A2 each equaling 50% (n ϭ 3). Differences between results with ␣ alone and with ␤1 coinjection were insignificant (unpaired Student's t test, p Ͼ 0.05). Consistent with the lack of effect on recovery from inactivation, after a 10 Hz train of 20 15-ms long pulses to Ϫ10 mV from a holding potential of Ϫ100 mV, current was reduced to 92.4 Ϯ 1.8% of its initial value with rH1 ␣ alone and to 91.9 Ϯ 1.9% with coinjection of ␤1 (n ϭ 7). Again, these values were not significantly different from each other (p Ͼ 0.05). Thus, in oocytes where effects of ␤1 subunit RNA coinjection have been verified by recordings of significantly increased Na ϩ current levels, there was no significant difference in Na ϩ current time course, voltage dependence of activation, voltage dependence of inactivation, or recovery from inactivation when comparing Na ϩ currents due to injection of rH1 Na ϩ channel ␣ subunit RNA alone or when coinjected with ␤1 subunit RNA. DISCUSSION Our results show that ␤1 subunit mRNA is expressed in cardiac muscle cells as assessed by high resolution in situ hybridization. Thus, ␤1 subunits are available to modulate rH1 Na ϩ channels in vivo. These studies complement previous work showing that ␤1 subunit mRNA and protein are present in the heart without defining the cell-type expressing them (Sutkowski and Catterall, 1990;Isom et al., 1992;Tong et al., 1993;Yang et al., 1993;Makita et al., 1994).
Coexpression of rH1 ␣ subunits and ␤1 subunits in Xenopus oocytes substantially increases the level of Na ϩ currents. Since both ␤1 subunits and ␣ subunits are present in cardiac myocytes, it is likely that ␤1 subunits associate with ␣ subunits and increase the expression of functional Na ϩ channels in cardiac cells as well. Purified Na ϩ channels from chicken and rat heart do not have associated ␤1 subunits (Lombet and Lazdunski, 1984;Cohen and Levitt, 1993). However, ␤1 subunits dissociate easily from detergent-solubilized Na ϩ channels ) and therefore may have been lost in purification. Since the purified preparations of Na ϩ channels from heart have not been functionally reconstituted, it remains to be determined whether these preparations which appear to lack ␤1 subunits can function in voltage-activated ion conductance.
Injection of cardiac (rH1) ␣ subunit mRNA into oocytes gave currents with comparable kinetics and voltage dependence to Na ϩ currents in native cardiac myocytes (Satin et al., 1992). Recordings from our laboratory in neonatal rat ventricular myocytes (Qu et al., 1994) gave half-activation values of Ϫ34 mV and half-inactivation values of Ϫ58 mV. Currents expressed in oocytes after injection of rH1 ␣ subunit RNA were half-activated at Ϫ28 mV and half-inactivated at Ϫ52 mV (see Fig. 5B), only 6 mV more positive than Na ϩ currents in ventricular myocytes. These small differences contrast sharply with injection of brain or muscle Na ϩ channel ␣ subunit RNA alone into Xenopus oocytes. Those currents exhibit abnormally slow kinetics and positively shifted voltage dependences of activation and inactivation which were normalized by coinjection of ␤1 subunits (Isom et al., 1992;Makita et al., 1994). Thus, ␤1 subunits have much more striking functional effects on brain and skeletal muscle Na ϩ channels than on cardiac Na ϩ channels.
During development of both rat retinal ganglion cells and rat forebrain neurons in vivo, expression of ␤ subunits and their assembly with ␣ subunits is concurrent with a 5-to 10-fold increase in the number of Na ϩ channels (Wollner et al., 1988;Scheinman et al., 1989;Sutkowski and Catterall, 1990). Thus, expression and assembly of ␤ subunits may be a rate-limiting step in Na ϩ channel expression. In addition, ␤1 subunits stabilize purified and reconstituted brain Na ϩ channels. Channel function was completely lost upon selective removal of ␤1 subunits . Our results showing that coinjection of ␤1 subunit mRNA significantly increases heart Na ϩ channel expression in Xenopus oocytes are consistent with the hypothesis that ␤1 subunits are important for the biosynthesis, assembly, and stabilization of the cardiac Na ϩ channel in vivo.