Isoform-Specific Modulation of Voltage-Gated Na+ Channels by Calmodulin
Calmodulin (CaM) is a calcium-sensing protein that binds to Na+ channels, with unknown functional consequences. Wild-type CaM produced a hyperpolarizing shift in the steady-state availability of expressed skeletal muscle (μ1) but not cardiac (hH1) Na+ channels. Mutant CaM1234 did not alter the voltage dependence or the kinetics of gating of either μ1 or hH1. Mutation of the highly conserved IQ motif in the carboxyl terminus of both isoforms (IQ/AA) slowed the kinetics of current decay and abolished the effect of wild-type CaM on μ1, but did not alter hH1 currents. The IQ/AA mutation eliminated CaM binding to the carboxyl terminus of both μ1 and hH1 channels. Inhibition of Ca2+/CaM kinase (CaM-K) slowed the current decay, the rate of entry into inactivation, and shifted the voltage dependence of hH1 in the depolarizing direction independent of CaM overexpression with no effect on μ1 Na+ channels. CaM signaling modulates Na+ currents in an isoform-specific manner, via direct interaction with skeletal muscle Na+ channels and through CaM-K in the case of the cardiac isoform. The full text of this article is available at http://www.circresaha.org.
Voltage-gated Na+ channels are essential for the generation of action potentials and cell excitability.1 Na+ channels are responsible for the upstroke of the action potential in excitable cells and are the target of local anesthetic drugs. Na+ channels are composed of a pore-forming α-subunit and smaller β-subunits that have been implicated in assembly, trafficking, and posttranslational modulation of the α-subunit.2
A number of interacting signaling pathways modulate the Na+ channel. The function of neuronal and cardiac Na+ channels is altered by cAMP-dependent protein kinase (PKA) phosphorylation of residues in the linker between domains I and II. PKA-dependent phosphorylation of the neuronal channel reduces current density,3,4⇓ but directionally opposite effects are seen with cardiac Na+ channels.5 The α-subunit of the skeletal muscle channel is not affected by PKA despite the presence of consensus phosphorylation sites.6,7⇓ In contrast to PKA, protein kinase C (PKC) alters the function of all mammalian sodium channel isoforms.8–13⇓⇓⇓⇓⇓ PKC reduces channel conductance and alters gating in an isoform-specific manner; an effect largely attributed to phosphorylation of a highly conserved serine residue in the linker between domains III to IV of the channel. There is also evidence that neuronal sodium channel gating is modulated by a G-protein–coupled pathway acting through the carboxyl terminus.14
Calmodulin (CaM) is a ubiquitous, small (16.7 kDa) calcium-binding protein that acts as a Ca2+ sensor translating changes in cytoplasmic Ca2+ into cellular responses by interacting with a diverse group of signaling molecules.15 A group of kinases, Ca2+/CaM-dependent kinases (CaM-K) are activated by CaM and phosphorylate a number of functionally diverse effectors from ion channels and receptors to transcriptional activators. Ion channels are prominent targets of CaM; for example, CaM is bound to L-type voltage–dependent Ca2+ channels and mediates Ca2+-dependent inactivation.16–19⇓⇓⇓ The possibility that CaM modulates the Na+ current was suggested by yeast 2-hybrid experiments demonstrating the interaction of voltage-dependent sodium channels and CaM through an IQ-like motif in the carboxyl terminus of the α-subunit.20
The functional consequences of CaM binding to Na+ channels are unknown. We demonstrate that coexpression of CaM with Na+ channels alters the function of the skeletal muscle (μ1) channels but not the cardiac (hH1) isoform. Further, mutation of the IQ motif in the carboxyl terminus of the μ1 skeletal muscle Na+ channel alters inactivation gating and eliminates the effect of CaM on the expressed current. Inhibition of CaM-K by KN-93, but not the CaM-KII–specific inhibitor autocamtide-2 peptide (AIP), slowed cardiac Na+ current decay, shifted the availability curve to more depolarized potentials, and slowed entry into inactivated states. Mutating the IQ motif eliminated CaM binding to carboxyl terminal peptides of both isoforms.
Materials and Methods
Transfections of HEK293 and C2C12 Cell Lines
Human embryonic kidney cells (HEK293; American Type Culture Collection, Manassas, Va) were used to express the cloned sodium channels μ1 and hH1 in presence of wild-type or mutant CaM1234.17 C2C12 cells (American Type Culture Collection) are a subclone derived from a mouse myoblast cell line expressing characteristic muscle proteins.21 Cells were grown in DMEM high glucose supplemented with FBS 10%, l-glutamine (2 mmol/L), penicillin (100 U/mL), and streptomycin (10 mg/mL) in a humidified 5% CO2 atmosphere. Transfections of the cells were performed using lipofectamine according to the manufacturer’s instructions. Cells were patch-clamped or stained 1 to 2 days after transfection.
Macroscopic Na+ currents from transfected cells were recorded using the whole-cell configuration of patch-clamp technique.22 Cells with currents no larger than 6 nA were clamped with an Axopatch 200 patch-clamp amplifier using low-resistance electrodes and typical series resistance compensation of >80% that minimized voltage-clamp errors. Membrane currents were filtered at 5 kHz and digitized with 12-bit resolution. In all experiments, recording was begun 10 minutes after whole-cell access was attained with the exception of the CaM-K inhibitor experiments where recording was begun 15 minutes after patch rupture to ensure stabilization of the time-dependent shift in gating.
Solutions and Reagents
The patch pipette contained (in mmol/L) 35 NaCl, 105 CsCl, and 10 Cs-HEPES (pH 7.3) with no Ca2+ buffering. All of the experiments were initially performed using 105 mmol/L CsF in the pipette; however, due to possible confounding effects on phosphorylation reactions and Ca2+ buffering, the experiments were repeated with the CsCl-containing intracellular solution. No differences in the response to CaM overexpression were observed with the different pipette solutions. All of the data reported in the text and in the tables were obtained using CsCl with the exception of experiments in C2C12 cells, where CsF data are reported.
For the experiments with 504 nmol/L free Ca2+, 5 mmol/L BAPTA and 3.7 mmol/L CaCl2 were added to the patch pipette. For the experiments with no free Ca2+, 10 mmol/L BAPTA was added to the pipette solution and 0.1 mmol/L EGTA was added to the bath solution. The bath solution contained (in mmol/L) 150 NaCl, 2 KCl, 2 CaCl2, 1 MgCl2, 10 glucose, and 10 Na-HEPES (pH 7.4). The CaM-K inhibitor KN-93 (Calbiochem) and the inactive analog, KN-92, were added to the patch pipette at concentrations of 10 μmol/L. KN-92 and KN-93 were dissolved in DMSO (10 mmol/L stock solution); the final concentration of DMSO was 0.1%, which had no significant effect on the currents (data not shown). The CaM-KII–specific inhibitor, AIP (Calbiochem) was added to the pipette solution at 100 nmol/L. Experiments were performed at room temperature, 22°C to 23°C.
Fusion proteins of the μ1 and hH1 carboxyl termini containing the IQ motif and the mutated motif IQ/AA were constructed in pGEX using the glutathione S-transferase (GST) Gene Fusion System (Pharmacia). The Na+ channel-specific sequences of the fusion peptides were μ1 (residues 1689 to 1742): ALKQTMEEKFMAANPSKVSYEPITTTLKRKQEEVCAIKIQRAYRRHLLQRSVKQ and hH1 (residues 1870 to 1923): ALKIQMEEKFMAANPSKISYEPITTTLRRKHEEVSAMVIQRAFRRHLLQRSLKH, where the underlined amino acids were changed to alanines in the IQ/AA mutants. Proteins were produced and purified according to the manufacturer’s instructions. Pull-down assays were performed by mixing 10 μg (1:1 molar ratio of NaIQ:CaM) or 30 μg (3:1 molar ratio of NaIQ:CaM) of each fusion protein with glutathione beads (70 μL). Binding of the beads to the GST-tagged fusion proteins was performed by incubating for 30 minutes at 4°C in solution A consisting of 25 mmol/L Tris, pH 7.6, and 1 mmol/L CaCl2. After a brief centrifugation, the supernatants were removed and the beads were mixed and incubated with solution A in the absence and presence of CaM for 2 hours at 4°C. The mixtures were washed and spun briefly. Then the proteins bound to the beads were eluted with SDS-Page 2X sample buffer (70 μL) and 30 μL of each sample was run on a 12.5% gel SDS-PAGE gel. After transfer to PVDF membranes CaM was detected by an anti-CaM monoclonal antibody (Upstate Biotechnology) and a horse radish peroxidase-linked secondary antibody.
Data are expressed as mean±SEM. A 2-tailed Student’s t test was used to assess statistical significance when appropriate. Differences were considered to be significant at a value of P<0.05.
Effects of CaM Coexpression With μ1 Skeletal Muscle Na+ Channels
To assess the effect of CaM binding on skeletal muscle sodium channels, we cotransfected HEK293 cells with μ1 and wild-type CaM or a mutant form of CaM with strongly impaired Ca2+ binding (CaM1234).17 In the absence of intracellular Ca2+ buffering, CaM overexpression did not significantly affect the whole-cell current (Figure 1A), current-voltage relationship (Figure 1B), voltage dependence of activation, time constants of inactivation (τh), recovery from inactivation, or the rate of entry into inactivation (Table 1). However, the steady-state availability relation was significantly shifted in the hyperpolarizing direction (Figure 1C). In contrast to wild-type CaM, coexpression of mutant CaM1234 with the μ1 Na+ channel produced no significant change in steady-state availability compared with the μ1 α-subunit expressed alone. The presence of the β1-subunit did not significantly alter the steady-state availability curve of the μ1 channel compared with the α-subunit alone, nor did it alter the hyperpolarizing shift produced by CaM. Table 1 summarizes the effects of wild-type and mutant CaM on the functional properties of the μ1 skeletal muscle Na+ channel expressed in the presence and absence of the β-subunit with CaM.
The action of CaM on effector molecules often requires Ca2+; therefore, we studied the effect of different free intracellular Ca2+ concentrations on the CaM modulation of μ1 sodium channels. Our standard recording conditions included no intracellular Ca2+ buffering and 2 mmol/L external Ca2+. There was no significant difference in the effect of CaM on μ1 Na+ current when [Ca2+]i was buffered to 504 nmol/L compared with control recording conditions (data not shown). However, when extracellular Ca2+ was buffered with 0.1 mmol/L EGTA and 10 mmol/L BAPTA was added to the intracellular solution to ensure the removal of free Ca2+, the CaM-induced shift of the steady-state inactivation curve was no longer present (Figure 1D). Thus, the effect of CaM on gating of the skeletal muscle Na+ current is Ca2+-dependent (Table 1).
Effects of CaM Coexpression on the hH1 Cardiac Na+ Current
The carboxyl terminal IQ-like binding motif is conserved across all isoforms of the mammalian Na+ channel. Therefore, we examined the effect of CaM on the function of the human heart sodium channel, hH1. When coexpressed with hH1, neither wild-type CaM nor mutant CaM1234 significantly altered any whole-cell property of the current (Table 2). Expression of wild-type CaM and mutant CaM1234 proteins was confirmed by immunofluorescence staining (data not shown). CaM coexpression had no effect on the hH1 Na+ current either in the absence or presence of the β1-subunit at any intracellular Ca2+ concentration. A cell line stably expressing hH1 was used for the β1-subunit coexpression experiments. An ≈−10 mV shift in the steady-state inactivation curve is observed in these cells compared with cells transiently transfected with hH1; however, CaM overexpression had no effect on the voltage dependence of inactivation. The rates of entry into and recovery from inactivation of the stably expressed channels was slower than transiently transfected hH1, but CaM overexpression had no significant effect on the kinetics of inactivation (Table 2).
Effects of CaM Binding on Skeletal Muscle Na+ Currents Recorded From C2C12 Cells
In order to study the effect of CaM on native skeletal muscle Na+ channels, we transfected wild-type CaM and mutant CaM1234 into C2C12 cells, a mouse myoblast cell line. C2C12 cells showed typical skeletal muscle sodium currents (Figure 2 and Table 1), as previously reported.21 The time constants of decay (τh) and recovery from inactivation are modestly but significantly slower in C2C12 cells compared with expressed μ1 currents recorded under identical conditions (Table 1). In the absence of exogenous CaM, the steady-state availability of C2C12 Na+ currents is shifted ≈−20 mV compared with expressed μ1 currents. Expression of wild-type CaM in C2C12 cells significantly shifts the steady-state availability of the Na+ currents in the hyperpolarizing direction (Figure 2C). However, coexpression of mutant CaM1234 did not shift the voltage dependence of the steady-state availability relationship (Figure 2C). Neither wild-type nor mutant CaM1234 altered the other whole-cell properties of C2C12 Na+ current (Figure 2, Table 1).
Mutation of CaM Binding Site
The isoleucine-glutamine (IQ) CaM binding domain is essential for CaM-mediated Ca2+-dependent inactivation of the L-type Ca2+ current.17,18⇓ Yeast 2-hybrid assays have established CaM binding to an IQ-like motif on the carboxyl terminal region of Na+ channels.20 In order to determine the role of the Na+ channel IQ motif in CaM-induced shifts in availability, we mutated IQ to alanines in both μ1 (I1727A-Q1728A, μ1-IQ/AA) and hH1 (I1908A-Q1909A, hH1-IQ/AA) Na+ channels and then coexpressed these mutant channels with wild-type CaM and mutant CaM1234. The IQ/AA mutations in the μ1 background significantly altered the whole-cell current compared with the wild-type current. The current decay became more obviously biexponential with the addition of a slow component of decay that resulted in a pedestal of current of ≈5% of the peak at 30 ms after the voltage step (Figure 3A). The peak IV relationship (Figure 3B) was not altered but the steady-state availability curve of μ1-IQ/AA was shifted ≈−10 mV compared with μ1-WT (Figure 3C, Table 1). Interestingly, by mutating the IQ motif, we abolished the CaM-induced shift of the steady-state availability relationship seen with coexpression of wild-type CaM and wild-type μ1 (Figure 3C, Table 1).
In contrast to μ1-IQ/AA, hH1-IQ/AA did not exhibit changes in whole-cell current decay compared with wild-type hH1. The wild-type hH1 current decay was best fitted with a biexponential function, and neither the time constants nor amplitudes of the components were altered by the IQ/AA mutations. Not surprisingly, coexpression of wild-type CaM with hH1-IQ/AA had no effect on any of the whole-cell properties of the current (Table 2).
CaM Binding to the IQ Motif
The colocalization of CaM and the Na+ channel is further supported by protein binding experiments. GST-fusion proteins of carboxyl-terminal regions containing the wild-type IQ motifs and the IQ/AA mutants of the hH1 and μ1 Na+ channels were made. The carboxyl terminal μ1 and hH1 fusion proteins with an intact IQ binding motif interact with CaM. In contrast, neither of the IQ/AA mutant proteins are able to pull-down CaM (Figure 4). Therefore, binding of CaM to the IQ motif appears to be necessary for the modulation of μ1 channel gating but is not sufficient in that binding to the hH1 carboxyl terminus does not alter gating of the cardiac isoform.
Effect of CaM-K Inhibition on Na+ Currents
The cellular effects of CaM often involve downstream signaling through CaM-K. The activity of Na+ channels is known to be modulated through different phosphorylation pathways. Because CaM binds to an IQ-like motif located in the C-terminus of Na+ channel, phosphorylation of the channel by CaM-K is possible. The isoform-specific differences in the response to CaM may result from differences in the channels as substrates for phosphorylation by CaM-K. To assess this possibility, we used a nonselective CaM-K inhibitor, KN-93, and its inactive analog, KN-92. We applied KN-93 and KN-92 intracellularly at a concentration of 10 μmol/L to μ1 Na+ channels expressed alone and in the presence of CaM.
KN-93 and the inactive analog KN-92 had no effect on the voltage dependence or kinetics of μ1 gating either in the presence or absence of CaM overexpression. Specifically, neither compound altered the rate of current decay (Figure 5B), rates of recovery from or entry into inactivated states (Table 1). Further, neither compound altered the CaM-induced hyperpolarizing shift of the steady-state availability relationship of μ1 Na+ channels (Figure 5A). The CaM-KII–specific inhibitor AIP (100 nmol/L) was applied to expressed μ1 currents with results similar to the KN compounds. AIP did not alter the voltage dependence or kinetics of gating of μ1 channels and did not diminish the hyperpolarizing shift of the steady-state inactivation relationship induced by CaM overexpression (Table 1).
In contrast, KN-93 had robust effects on hH1 current, independent of CaM coexpression. KN-93 slowed the current decay with the most prominent effect on the fast time constant (Figure 6A, Table 2), shifted the steady-state inactivation relationship in depolarizing direction (Figure 6B), and slowed entry into inactivated states (Figure 6D, Table 2).
KN-92 also had effects on hH1, but not μ1 currents: it shifted the steady-state inactivation curve in the hyperpolarizing direction and produced an apparent slowing of recovery at −120 mV. However, the currents are not fully reprimed at this voltage in the presence of KN-92. KN-92, and to a lesser extent KN-93, retards recovery from inactivation compared with hH1 alone; therefore, KN-induced slowed recovery from inactivation cannot be considered a specific effect of CaM-K inhibition. The slowing of recovery by these compounds was not observed in expressed μ1 channels.
In contrast, AIP, a selective CaM-KII inhibitor, had no effect on channel activity and did not block the depolarizing shift in the steady-state inactivation curve by the subsequent addition of KN-93 to the bath (data not shown). Thus, a CaM-K other than CaM-KII, possibly CaM-KIV, which is richly expressed in the heart, may modulate the shift in the voltage dependence of inactivation of cardiac Na+ channels.
CaM is the most ubiquitous Ca2+-sensing protein in mammalian cells. The Ca2+/CaM complex directly modulates the function of a number of proteins including a family of serine/threonine kinases (CaM-K), ion channels, and receptors. CaM that is constitutively tethered to L-type Ca2+ channels mediates Ca2+-dependent inactivation and facilitation of the current.16–18⇓⇓ On Ca2+ binding, CaM associates with a conserved IQ motif23 in the carboxyl terminus of the channel. The presence of a highly conserved IQ-like motif in the carboxyl terminus of the Na+ channel suggested the possibility of CaM-mediated modulation of Na+ currents. Supportive evidence for such a mechanism included the demonstration of Ca2+-dependent and Ca2+-independent binding of CaM to a carboxyl terminal fragment of the Na+ channel containing the IQ motif.20 The functional consequences, if any, of CaM binding to the Na+ channel were previously unknown.
We demonstrate that CaM does indeed modify the function of the μ1 skeletal muscle isoform of the Na+ channel, shifting the steady-state inactivation curve in the hyperpolarizing direction (Figure 1, Table 1). The CaM effect on gating requires intact Ca2+ binding because a mutant, CaM1234, with impaired Ca2+ binding does not alter the gating of the μ1 Na+ channel. In addition, buffering of Ca2+ eliminates the wild-type CaM-mediated shift in steady-state availability (Figure 1D, Table 1). As expected, buffering Ca2+ produced a hyperpolarizing shift in channel gating, and we cannot exclude the possibility that this shift masks the CaM effect. CaM modulation of the skeletal muscle Na+ channel does not require the presence of the β1-subunit, nor does β1 alter the effect of CaM on the expressed current (Table 1). The modulation of the expressed μ1 current was recapitulated in C2C12 cells, a bona fide skeletal muscle cell line. These data suggest that native skeletal muscle channels, regardless of subunit composition, are targets for CaM.
Analogous to the Ca2+ channel, an IQ-like motif is a crucial component of the functional effects of CaM on skeletal muscle Na+ channels. Mutation of the IQ motif eliminates CaM binding to (Figure 4) and modulation (Figure 3) of μ1 skeletal muscle currents. Our data indicate that the IQ motif has important functional effects on skeletal muscle Na+ currents independent of CaM binding. Indeed, the μ1-IQ/AA mutation itself slows the decay of the whole-cell current and shifts the steady-state inactivation curve to hyperpolarized potentials relative to wild-type μ1. In fact, the shift in the availability curve produced by mutating IQ to AA exceeds the shift produced by coexpression of CaM with wild-type μ1. It is unlikely that the effect of the IQ to AA mutation is simply the result of a loss of CaM binding because the direction of the shift is opposite to what would be predicted by the elimination of CaM binding (Figure 3C). Instead, this suggests that the IQ region of the skeletal muscle Na+ channel is an important component of the channel’s inactivation machinery that is modulated by CaM-mediated signaling.
IQ-mediated binding of CaM to Na+ channels is a necessary but not sufficient condition for modulation of channel function. hH1 and μ1 Na+ channels have IQ motifs with similar CaM binding capacity (Figure 4), yet CaM coexpression does not alter the function of hH1 channels. The difference in the functional effects of CaM on the Na+ channel isoforms may be the result of the conformation of CaM bound to the channel. It has been shown that CaM can bind to the IQ motif of the rat brain II Na+ channel in its Ca2+-bound and Ca2+-free (apo-CaM) states.20 Indeed, Ca2+ buffering eliminates CaM modulation of μ1 channel gating (Figure 1D). However, the functional roles of the IQ motifs in skeletal muscle and cardiac Na+ channels appear to differ independent of CaM binding, as revealed by the IQ/AA mutant channels. Unlike the μ1 mutant channel, hH1-IQ/AA does not alter gating (Tables 1 and 2⇑).
Alternatively, binding of CaM to the Na+ channel may serve to facilitate phosphorylation of the channel complex by CaM-K. We do not have direct evidence for phosphorylation of either channel isoform; however, inhibition of CaM-K (possibly CaM-KIV) modifies cardiac Na+ channel inactivation, shifting the steady-state inactivation curve in the depolarizing direction, slowing entry into inactivated states, and hastening recovery when compared with the inactive analog KN-92. KN-92 itself has effects on the cardiac Na+ channel (shifting the inactivation curve in the hyperpolarizing direction and slowing recovery) that cannot be ascribed to CaM-K inhibition. The differential effects of CaM-K inhibition on skeletal muscle and cardiac channels suggest the possibility of distinct functionally significant phosphorylation sites on each of the channel isoforms. However, the existence and precise identity of such sites remain to be demonstrated.
CaM signaling appears to fine-tune the function of voltage-dependent Na+ channels in isoform-specific ways. The physiological significance of such modulation is uncertain, but recruitment of CaM signaling tends to stabilize inactivation of skeletal muscle and cardiac channels, albeit by distinct molecular mechanisms.
Interestingly, the C-terminal region of the cardiac Na+ channel is a hot spot for mutations that cause inherited cardiac arrhythmias, such as long-QT and Brugada syndromes.24–30⇓⇓⇓⇓⇓⇓ We have recently demonstrated that the C-terminus of the Na+ channel plays a role in fast inactivation.31 A common theme is that these disease-causing mutations alter gating and often disrupt inactivation of the Na+ channel. The mutations may directly alter Na+ channel function; however, it is possible that the effect of the mutation is through interruption of CaM/CaM-kinase signaling leading to destabilization of inactivation of the Na+ channel.
In the heart under pathophysiological conditions associated with increased intracellular Ca2+, CaM-KII appears to enhance the arrhythmogenic transient inward current.32 Our data suggest that CaM may also regulate activation and repolarization of heart cells through effects on Na+ channels altering the susceptibility to arrhythmias. Similarly, modulation of Na+ channel availability may regulate the frequency of action potential firing in skeletal muscle, a possible mechanistic link to myotonias.
This work was supported by grants from the National Institutes of Health R01 HL-50411 (G.F.T.), R01 HL-52768 (E.M.), and SCOR grant P50 HL-52307, (E.M., D.T.Y., G.F.T.). I.D. is a research fellow of the Heart and Stroke Foundation of Canada. E.M. holds the Michel Mirowski, MD, Professorship of Cardiology of the Johns Hopkins University.
This manuscript was sent to Harry A. Fozzard, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
Original received December 21, 2001; revision received January 24, 2002; accepted January 28, 2002.
- ↵Catterall WA. Cellular and molecular biology of voltage-gated sodium channels. Physiol Rev. 1992; 72: S15–S48.
- ↵Rossie S, Catterall WA. Cyclic-AMP–dependent phosphorylation of voltage-sensitive sodium channels in primary cultures of rat brain neurons. J Biol Chem. 1987; 262: 12735–12744.
- ↵Rossie S, Catterall WA. Phosphorylation of the α subunit of rat brain sodium channels by cAMP-dependent protein kinase at a new site containing Ser686 and Ser687. J Biol Chem. 1989; 264: 14220–14224.
- ↵Murphy BJ, Rogers J, Perdichizzi AP, Colvin AA, Catterall WA. cAMP-dependent phosphorylation of two sites in the α subunit of the cardiac sodium channel. J Biol Chem. 1996; 271: 28837–28843.
- ↵West JW, Numann R, Murphy BJ, Scheuer T, Catterall WA. A phosphorylation site in the Na+ channel required for modulation by protein kinase C. Science. 1991; 254: 866–868.
- ↵Numann R, Catterall WA, Scheuer T. Functional modulation of brain sodium channels by protein kinase C phosphorylation. Science. 1991; 254: 115–118.
- ↵Bendahhou S, Cummins TR, Potts JF, Tong J, Agnew WS. Serine-1321–independent regulation of the μ1 adult skeletal muscle Na+ channel by protein kinase C. Proc Natl Acad Sci U S A. 1995; 92: 12003–12007.
- ↵Qu Y, Rogers JC, Tanada TN, Catterall WA, Scheuer T. Phosphorylation of S1505 in the cardiac Na+ channel inactivation gate is required for modulation by protein kinase C. J Gen Physiol. 1996; 108: 375–379.
- ↵Murray KT, Hu NN, Daw JR, Shin H-G, Watson MT, Mashburn AB, George AL Jr. Functional effects of protein kinase C activation on the human cardiac Na+ channel. Circ Res. 1997; 80: 370–376.
- ↵Ma JY, Li M, Catterall WA, Scheuer T. Modulation of brain Na+ channels by a G-protein–coupled pathway. Proc Natl Acad Sci U S A. 1994; 91: 12351–12355.
- ↵Qin N, Olcese R, Bransby M, Lin T, Birnbaumer L. Ca2+-induced inhibition of the cardiac Ca2+ channel depends on calmodulin. Proc Natl Acad Sci U S A. 1999; 96: 2435–2438.
- ↵Rhoads AR, Friedberg F. Sequence motifs for calmodulin recognition. FASEB J. 1997; 11: 331–340.
- ↵Abriel H, Wehrens XH, Benhorin J, Kerem B, Kass RS. Molecular pharmacology of the sodium channel mutation D1790G linked to the long-QT syndrome. Circulation. 2000; 102: 921–925.
- ↵An RH, Wang XL, Kerem B, Benhorin J, Medina A, Goldmit M, Kass RS. Novel LQT-3 mutation affects Na+ channel activity through interactions between α- and β1-subunits. Circ Res. 1998; 83: 141–146.
- ↵Benhorin J, Taub R, Goldmit M, Kerem B, Kass RS, Windman I, Medina A. Effects of flecainide in patients with new SCN5A mutation: mutation-specific therapy for long-QT syndrome? Circulation. 2000; 101: 1698–1706.
- ↵Bezzina C, Veldkamp MW, van Den Berg MP, Postma AV, Rook MB, Viersma JW, van Langen IM, Tan-Sindhunata G, Bink-Boelkens MT, van Der Hout AH, Mannens MM, Wilde AA. A single Na+ channel mutation causing both long-QT and Brugada syndromes. Circ Res. 1999; 85: 1206–1213.
- ↵Wei J, Wang DW, Alings M, Fish F, Wathen M, Roden DM, George AL Jr. Congenital long-QT syndrome caused by a novel mutation in a conserved acidic domain of the cardiac Na+ channel. Circulation. 1999; 99: 3165–3171.
- ↵Wu Y, Roden DM, Anderson ME. Calmodulin kinase inhibition prevents development of the arrhythmogenic transient inward current. Circ Res. 1999; 84: 906–912.