Mechanisms of a Human Skeletal Myotonia Produced by Mutation in the C-Terminus of NaV1.4: Is Ca2+ Regulation Defective?

Mutations in the cytoplasmic tail (CT) of voltage gated sodium channels cause a spectrum of inherited diseases of cellular excitability, yet to date only one mutation in the CT of the human skeletal muscle voltage gated sodium channel (hNaV1.4F1705I) has been linked to cold aggravated myotonia. The functional effects of altered regulation of hNaV1.4F1705I are incompletely understood. The location of the hNaV1.4F1705I in the CT prompted us to examine the role of Ca2+ and calmodulin (CaM) regulation in the manifestations of myotonia. To study Na channel related mechanisms of myotonia we exploited the differences in rat and human NaV1.4 channel regulation by Ca2+ and CaM. hNaV1.4F1705I inactivation gating is Ca2+-sensitive compared to wild type hNaV1.4 which is Ca2+ insensitive and the mutant channel exhibits a depolarizing shift of the V1/2 of inactivation with CaM over expression. In contrast the same mutation in the rNaV1.4 channel background (rNaV1.4F1698I) eliminates Ca2+ sensitivity of gating without affecting the CaM over expression induced hyperpolarizing shift in steady-state inactivation. The differences in the Ca2+ sensitivity of gating between wild type and mutant human and rat NaV1.4 channels are in part mediated by a divergence in the amino acid sequence in the EF hand like (EFL) region of the CT. Thus the composition of the EFL region contributes to the species differences in Ca2+/CaM regulation of the mutant channels that produce myotonia. The myotonia mutation F1705I slows INa decay in a Ca2+-sensitive fashion. The combination of the altered voltage dependence and kinetics of INa decay contribute to the myotonic phenotype and may involve the Ca2+-sensing apparatus in the CT of NaV1.4.


Introduction
Precise and coordinated activity of skeletal muscle results from highly regulated signals generated by the orchestrated activities of different ion channels. A spectrum of muscle disorders are caused by the mutations in different ion channels [1]. A number of mutations in different regions of the human voltage-gated Na channel, hNa V 1.4 have been reported to cause skeletal muscle disorders [2]. Mutations in the transmembrane domains and linker regions cause cold aggravated myotonia [3,4,5,6] yet only one case of cold aggravated myotonia [7], has been linked to a mutation in the CT of the skeletal muscle sodium channel, hNa V 1.4 (F1705I, Figure 1A) and functionally studied. The other mutation in the CT-Na V 1.4, E1702K has been linked to paramyotonia congenita [8], but not functionally studied. Interestingly, another mutation Q1633E, causes potassium aggravated myotonia and is located in the EF hand like (EFL) region, that is the region in and around Helix 1, and the loop between Helix 1 and Helix 2 of the CT-hNa V 1.4, and about 71 amino acids upstream of the E1705I mutation. Q1633E and F1705I have comparable electrophysiological effects including disruption of fast inactivation, slowed current decay, and a depolarized shift in the voltage dependence of availability [9]. hNa V 1.4 mutations that cause myotonic disorders are associated with changes in the kinetics and voltage dependence of gating generally resulting in a gain-of-function.
There is substantial evidence that the CT of Na V channels regulate the kinetics and voltage dependence of inactivation [10,11,12,13]. Moreover, Ca 2+ and CaM/CaM kinase (CaMK) distinctly modulate inactivation of different isoforms of Na V channels through interaction with structural motifs in the CT although the mechanisms are not fully understood [10,11,14,15,16,17,18]. There is evidence, particularly in HEK cells, that over expression of CaM will alter Na V 1.4 channel gating [10,15,16], and this is not unique to this channel but it also observed for other voltage dependent channels [15,19,20]. An EFL sequence [21,22] in the CT of Na V 1.5 and Na V 1.1 has been shown to influence Ca 2+ regulation of channel gating [14,17,18,23,24,25]. Other acidic residues in the H1-H2 loop of the proximal C-terminus may affect Ca 2+ sensitivity of channel gating as well [24,26], and the NMR solution and crystal structures in this region reveals binding of Ca 2+ [24] or Mg 2+ [27]. In addition to the EFL, sites in the III-IV linker are involved in Ca 2+ /CaM mediated regulation of gating [28,29,30]. Similar Ca 2+ /CaM sites are also present in the CT of Na V 1.4; however, their roles in channel regulation and relevance to disease causing mutations of the CT-Na V 1.4 are not known. The goals of this study are to understand the pathogenesis of temperature sensitive myotonia caused by the F1705I mutation, and to use this naturally occurring mutation (and the species specific differences) to better understand the regulation of the Na V channels by Ca 2+ . We demonstrate that the hNa V 1.4 F1705I alters the voltage dependence of inactivation and the temperature sensitivity of current kinetics. In addition, key residues in the EFL alter the CaM and Ca 2+ dependence of channel gating which may also contribute to the myotonia phenotype.

Plasmid Construction
The EYFP fused channel construct Na V 1.4-EYFP was prepared as described previously [15]. The myotonia causing mutation in the rat CT, rNa V

Transfection of Cells
Approximately 0.75610 6 Human embryonic kidney cells (HEK293; American Type Culture Collection, Manassas, VA) were cultured in 6-well tissue culture dishes in DMEM supplemented with 10% FBS, L-glutamine (2 mmol/L), penicillin (A) A schematic of the structured region of the C-terminus of hNa V 1.4 between amino acids residues 1788 and 2040, the predicted helices are labeled H1-H6. The location of the EFL residues in and around H1 harbors species specific variations in the key Ca 2+ sensing residues in hNa V 1.4 (G1613S and A1636D) compared with the rat isoform.

Electrophysiology
HEK293 cells expressing wild type or mutant EYFP-tagged Na V 1.4 channels were selected for recording. In some experiments, cells expressing both yellow and cyan fluorophores that were co-transfected with tagged Na V 1.4 channel variants and CaM or CaM 1234 were selected for patching. Cells were patch clamped with an Axopatch 200B patch-clamp amplifier using pipettes with tip resistances of 1-3 MV and typical series resistance compensation of .90% to minimize voltage clamp errors. Current recording was initiated 10 minutes after establishing whole-cell configuration, and currents were filtered at 5 kHz.
All the solutions used in this study were prepared as described previously [14,17,31,32]. The bath solution contained (in mmol/ L): 145 NaCl, 4 KCl, 1.8 CaCl 2 , 1 MgCl 2 , 10 glucose and 10 Na-HEPES (pH 7.4). The Ca 2+ free patch pipette solution contained (in mmol/L): 10 NaF, 100 CsF, 20 CsCl 2 , 20 BAPTA, 0 CaCl 2 and 10 HEPES, pH adjusted to 7.35 with CsOH. The 0.5 mmol/L Ca 2+ patch pipette solution contained (in mmol/L): 10 NaF, 100 CsF, 20 CsCl 2 , 5 BAPTA, 4 CaCl 2 and 10 HEPES, pH adjusted to 7.35 with CsOH. The osmolarity of the bath and pipette solutions were equalized using glucose. The free [Ca 2+ ] in the solutions was estimated with WEBMAX Standard software (http://www.stanford.edu/,cpatton/webmaxcS.htm) and found to be about 0.5 mM. The free [Ca 2+ ] in the solutions was verified by measurement using a Kwik-Tip Calcium ion-selective electrode (WPI) ( Figure S1 and Table S1 in File S1). We studied the effect of intracellular Ca 2+ on the voltage dependence of inactivation gating of rNa V 1.4 with chloride substituted for fluoride in the pipette solution and found no significant difference in the Ca 2+ sensitivity of gating ( Figure S2 in File S1).
Standard two-pulse protocols were used to generate the steadystate inactivation curves. The voltage dependence of steady-state fast inactivation was studied using 500 ms inactivation pre-pulses over a voltage range from 2140 to +30 mV in steps of 5 mV, followed by a 50 ms test pulse at 220 mV. Currents were normalized to the maximal current (I max ) and fit to a Boltzmann function of the form (y = [(A 1 2A 2 )/(1+e (x2x0)/dx) )]+A 2 ) to determine the membrane potential eliciting half-maximal inactivation (V 1/2 ), where A 1 and A 2 are maximum and minimum availabil- ities, respectively, x0 is equivalent to V 1/2 , and dx represents the slope factor. Activation curves were generated from a family of 50 ms test pulses from 2100 mV to +80 mV in 5 mV increments from a holding potential of 2120 mV. Peak currents at each membrane potential normalized to I max, were plotted to generate the I-V curves. Conductance (G) was calculated for peak current (I Peak ) at each membrane potential (V m ) using the equation (G = I Peak /(V m 2V Reversal )). G was normalized to G Max and fitted to a Boltzmann distribution to determine the membrane potential eliciting half-maximal activation (V 1/2 ). Deactivation was assessed using tail currents elicited by a test pulse of 0.5 ms to +40 mV followed by repolarization to a family of voltages ranging from 2180 to 250 mV. I Na decay rates were assessed by fitting to a single exponential decay ( Figure S3 in File S1). Recovery from inactivation was assessed by a standard two-pulse protocol with a first pulse duration of 30 ms and a second pulse of 30 ms to 220 mV with varying inter pulse intervals from 1 to 200 ms at 2120 mV. Exponential functions of the form y = y0+Ae 2x/t were fitted to recovery data to determine time constants (t rec ), where y0 is the offset and A is amplitude. Significance was assessed using unpaired student's t-test (Microcal Origin, Microcal Software Inc. MA), and p,0.05 was considered significant.

hNa V 1.4 F1705I Alters Current Kinetics and Gating
A mutation (F1705I) in the structured portion of the CT of hNa V 1.4 has been associated with cold aggravated myotonia ( Figure 1A). We studied the effect of hNa V 1.4 F1705I on channel function by transient expression in HEK293 cells at RT. Currents through wild type hNa V 1.4 and hNa V 1.4 F1705I channels were elicited by a family of depolarizing pulses ranging from 2100 mV to +80 mV from a holding potential of 2120 mV (Inset Figure 1B). The normalized peak current-voltage (I-V) relationships of wild type hNa V 1.4 and hNa V 1.4 F11705I were not different in 0.5 mM [Ca 2+ ] i mimicking the average intracellular [Ca 2+ ] in muscle ( Figure 1B and 1C). The voltage dependence (V 1/2 ) of activation of the wild type and mutant channels did not differ in the presence of 0.5 mM [Ca 2+ ] i ( Figure 1E and Table 1).
The voltage dependence of steady-state inactivation was altered by hNa V Figure 1E and Table 1). In the absence of a significant shift in the activation curve this produces a significant increase in the window current of hNa V Figure 1C); and the V 1/2 s of the activation (G2V) curves were unchanged in the absence of Ca 2+ ( Figure 1E and Table 1 Table 2). We studied the effect of hNa V 1.4 F1705I on current decay at 37uC, as patients with the mutation do not exhibit myotonia at normal body temperature. In contrast to RT, at 37uC in 0.5 mM [Ca 2+ ] i , the current decay of hNa V 1.4 F1705I is markedly hastened ( Figure 2C) and is not significantly different from wild type except at 230 mV ( Figures 2C, and 2D, Table 2). At 37uC the V 1/2 of steady-state inactivation of hNa V 1.4 F1705I was significantly shifted in hyperpolarizing direction compared to RT (V 1/2 at 37uC: 271.660.1 mV and V 1/2 at RT: 262.760.1 mV, p,0.001; Figure 2E). I Na decay, particularly at negative voltages, results from both inactivation and deactivation. We found no significant difference in the rates of deactivation at voltages between 2180 and 2100 mV (p,0.05; Figure S3 in File S1) similar to previous results [33]. However, at voltages from 290 mV to 260 mV, t Deactivation of hNa V 1.4 F1705I in the Ca 2+ free condition is significantly larger than in the wild type channel, with or without Ca 2+ (p,0.05; Figure S3 in File S1).
Our previous work showed that the cardiac isoform hNa V 1.5 is [Ca 2+ ] i sensitive, and lowering intracellular Ca 2+ shifts the V 1/2 of steady-state fast inactivation of hNa V 1.5 in hyperpolarizing direction [14]. However, hNa V 1.4 was insensitive to [Ca 2+ ] i and altering [Ca 2+ ] i does not significantly affect the voltage dependence of steady-state inactivation of wild type hNa V 1.4 channels ( Figure 1E). In contrast, the V 1/2 of steady-state inactivation of hNa V 1.4 F1705I was shifted significantly (,5 mV) in the depolarizing direction in the absence of intracellular Ca 2+

hNa V 1.4 F1705I Alters CaM Modulation
The F1705I mutation which is remote from the EFL region modifies the Ca 2+ sensitivity of gating. F1705 is predicted to be in Helix 5 of the CT hNa V 1.4 and is closer in the linear amino acid sequence to the CaM binding IQ motif in Helix 6 ( Figure 1A). Previously we have shown that CaM binds to the IQ motif and shifts the voltage dependence of inactivation of rat Na V 1.4 (rNa V 1.4) and Ca 2+ binding-deficient hNa V 1.5 channels that are mutated in the EFL [10,14,15]. We tested the hypothesis that the large changes in Na + current inactivation exhibited by hNa V 1.4 F1705I compared to wild type hNa V 1.4 are due to functional alterations in CaM interaction with the IQ motif. We examined the effects of CaM over expression on hNa V 1.4 F1705I and wild type hNa V 1.4 gating. As Ca 2+ can bind to CaM and modulate channel gating, we also studied effect of Ca 2+ binding deficient CaM or apo-CaM (CaM 1234 ) to delineate direct effects of Ca 2+ on the channel compared to Ca 2+ effects through CaM. In contrast to the wild type rat Na V 1.4 channel [15], over expression of both CaM and CaM 1234 with wild type human Na V 1.4 shifts the channel's I-V relation in the depolarizing direction in 0.5 mM [Ca 2+ ] i ( Figure 3A and Table 1). Co-expression of CaM and CaM 1234 with hNa V 1.4 significantly shifts the voltage dependence of activation in the depolarizing direction ( Figure 3D). In contrast, neither the I-V relationships or voltage dependence of activation of hNa V 1.4 F1705I were affected by co-expression with CaM or CaM 1234 (Figures 3B, 3E and Table 1).
We previously demonstrated that CaM is tethered to the CT of Na V 1.4, and CaM binding to the IQ motif shifts the steady-state inactivation of the rat isoform in a hyperpolarizing direction [10,15] (Table 1). In contrast, co-expression of CaM with hNa V 1.4 significantly shifts wild type channel availability in the depolarizing direction in 0.5 mM [Ca 2+ ] i (Figures 3C and D) and in the hyperpolarizing direction in the Ca 2+ free intracellular condition, when compared with the absence of exogenous CaM (Table 1; p,0.05). CaM 1234 over expression in 0.5 mM [Ca 2+ ] i shifts the V 1/2 of inactivation of hNa V 1.4 in the depolarizing direction compared to the absence of CaM 1234 ( Figure 3D; p,0 We speculate that the difference in the voltage dependence of gating of wild type human and rat Na V 1.4 channels when coexpressed with Ca 2+ -CaM/apo-CaM [15] may provide insights into the mechanisms of the Ca 2+ modulation of channel function in the myotonia mutation and in wild type channels.

Effect of rNa V 1.4 F1698I on Ca 2+ /CaM Regulation of Channel Gating
The wild type rat and human isoforms of Na V 1.4 exhibit distinct differences in gating and regulation by Ca 2+ and CaM. The orthologous mutation of hNa V 1.4 F1705I in the rat is rNa V 1.4 F1698I . The activation curve of rNa V 1.4 F1698I is significantly (p,0.05) shifted in the depolarizing direction (,+12 mV) compared to wild type rNa V 1.4 in 0.5 mM [Ca 2+ ] i ( Figures 4A, 4B and Table 1). In contrast to rNa V 1.4 F1698I , no activation shift was seen with the F1705I mutation in human Na V 1.4 ( Figure 1E). Similar shifts in the activation curves of other Na V 1.4 mutants that cause myotonia have previously been reported [34,35,36].
We next examined whether Ca 2+ altered rNa V 1.4 F1698I gating. The V 1/2 s of the activation of both the wild type rNa V Table 1). Thus the Ca 2+ sensitivity of steady-state inactivation of rNa V 1.4 was eliminated with a mutation remote in the linear amino acid sequence from the EFL motif. Although the rNa V 1.4 F1698I results in loss of Ca 2+ sensitivity of the channel, it also dramatically affects steady-state inactivation of the channel producing approximately ,+15 mV shift of the V 1/2 compared to   Table 1; p,0.05).

IQ-CaM Interaction is Unaffected by rNa V 1.4 F1698I
The significant alteration in Na + current properties demonstrated by rNa V 1.4 F1698I suggests the possibility of a disruption of CaM interaction with the IQ motif which is in vicinity of F1698. CaM or CaM 1234 co-expression does not alter the current kinetics ( Figure 5A) or the voltage dependence of activation of rNa V 1.4 F1698I ( Figure 5C). Similarly, CaM and CaM 1234 do not affect the activation of wild type rNav1.4 (Table 1). However, coexpression of CaM with rNa V 1.4 F1698I significantly shifts channel availability in the hyperpolarizing direction compared to the absence of CaM co-expression ( Figures 5B and 5C; p,0.05). CaM 1234 has no effect on the voltage dependence of steady-state inactivation in 0.5 mM [Ca 2+ ] i ( Figure 5C). CaM and CaM 1234 also have similar effects on wild type rNav1.4 availability ( Table 1). The time constants of recovery from inactivated states of rNa V 1.4 F1698I are unaffected by over expression of CaM or CaM 1234 (Table 1). In both species the orthologous myotonia mutations in the CT affect the Ca 2+ regulation of channel gating. In the rat channel rNa V 1.4 F1698I abolished the Ca 2+ sensitivity observed in the wild type rNa V 1.4. In contrast, hNa V 1.4 F1705I imparts sensitivity to Ca 2+ which is absent in wild type hNa V 1.4. We postulated that differences in the EFL region of the two channels contribute to the difference in Ca 2+ response of the orthologous myotonia mutations in the human and rat channels.

EFL Residues Mediate the Differences in Ca 2+ Regulation
We compared the amino acid sequences in the CT of the rat and human Na V 1.4 channels in the EFL and IQ motifs. In the region that includes the EFL (Figure 1A sensing through the EFL motif. We generated a triple mutant human channel, containing F1705I and replacing the nonconserved residues in the human EFL with the corresponding amino acids from the rat sequence hNa V 1.4 F1705I,G1613S,A1636D (hNa V 1.4 F1705I+GS/AD ). hNa V 1.4 F1705I+GS/AD and hNa V 1.4 have comparable peak current I-V relationships, and in both cases the voltage dependences of activation were not altered by the absence or presence of 0.5 mM Ca 2+ (Figures 6B, 6C and Table 1). Similar to rNa V 1.4 F1698I , steady-state inactivation of hNa V 1.4 F1705I+GS/AD was insensitive to changes in the [Ca 2+ ] i ( Figure 6D and Table 1).
Thus key residues in the EFL region contributed to the species differences in Ca 2+ sensitivity of mutations in Na V 1.4 that produce myotonia.
Remarkably hNa V 1.4 F1705I+GS/AD exhibits CaM regulation that recapitulates that of rNa V 1.4 F1698I . Over expression of CaM with hNa V 1.4 F1705I+GS/AD produces a hyperpolarizing shift in V 1/2 of steady-state inactivation compared to the mutant channel in the absence of CaM over expression ( Figures 6E, 6F and Table 1; p,0.05). Similar to rNa V 1.4 F1698I , CaM 1234 over expression with hNa V 1.4 F1705I+GS/AD does not alter the voltage dependence of steady-state inactivation compared to the mutant channel without CaM 1234 over expression ( Figure 6F and Table 1). The orthologous myotonia mutation on the rat channel background exhibits changes in Ca 2+ /CaM regulation that are completely different from that of hNa V 1.4 F1705I . However substituting the EFL residues present in the rat channel into the hNa V 1.4 F1705I background is sufficient to recapitulate the Ca 2+ /CaM regulation exhibited by rNa V 1.4 F1698I (Figures 6 and Table 1). This holds true for wild type channels, as mutation of the non-conserved amino acids of the wild type hNa V 1.4 EFL to match that of the rat sequence (G1613S and A1636D) reestablished Ca 2+ and CaM sensitivity of inactivation gating mimicking that of wild type rat Na V 1.4 ( Figures 6G and 6H, Table 1). Additionally, changing the EFL residues (G and A) from human channel into the wild type rat Na V 1.4 (rNa V 1.4 SG/DA ) or mutant rNa V 1.4 F1698I (rNa V 1.4-F1698I+SG/DA ) background restored Ca 2+ /CaM regulation similar to that displayed by hNa V 1.4 or hNa V 1.4 F1705I respectively (Table 1). Thus the differences in the key amino acids in the EFL region are associated with species-specific differences in Ca 2+ /CaM regulation of the wild type and mutant, human and rat Na V 1.4 channels.

Discussion
Myotonic mutations in hNa V 1.4 have been reported to affect Na + current inactivation properties. Wu et al. demonstrated a destabilization of fast inactivation without significant changes in activation or slow inactivation by the F1705I mutation [7]. However, the mechanism of the change in gating is uncertain as is the role of alteration of Ca 2+ or CaM modulation of the function of the mutant channel. We have demonstrated that the CT myotonia mutation, hNa V 1.4 F1705I slows the I Na decay, depolarizes the voltage dependence of inactivation and augments the window current, effects that may contribute to the cold-induced myotonic phenotype. Notably, hNa V 1.4 F1705I imparts intracellular Ca 2+ sensitivity to inactivation gating, a feature that is distinct from the wild type hNa V 1.4. However, the direction of the shift of steady-state inactivation of hNa V 1.4 F1705I channels is opposite to the direction of the Ca 2+ -induced shift observed with cardiac channel, hNa V 1.5 [14]. Our data suggest that two key residues in the EFL ( Figure 6, Table 1) modify gating of wild type hNa V 1.4 and the myotonia mutant hNa V 1.4 F1705I .
Sodium channel mediated myotonia is characterized electrophysiologically by a delay in inactivation that predisposes to repetitive depolarization and contraction of skeletal muscle after a brief stimulus. Previous studies suggested that a defect in fast inactivation, as in the CT mutant F1705I mutation, was sufficient to produce myotonia [7,37]. Our data shows, that along with a depolarizing shift in steady-state inactivation and slowed I Na decay, reduced intracellular Ca 2+ levels exaggerate both the depolarizing shift of inactivation ( Figures 1D, 1E and 1G) and the slowing of I Na decay of the current (Figures 2A and 2B). It appears that the largest effect of the F1705I mutation is on the voltage dependence and temperature sensitivity of gating but our data suggest that altered [Ca 2+ ] i sensitivity of the mutant channel potentiates symptoms of myotonia at low temperature. The experiments were designed to test the effects of Ca 2+ on channel gating and we explored the extremes of the range of concentrations. We do not believe that bulk Ca 2+ levels ever reach these levels but sub cellular distribution of ion concentrations are heterogeneous and local changes in Ca 2+ in this range are feasible [38].
Slowing of I Na decay, increased window current, altered voltage dependence of inactivation, along with disruption of Ca 2+ sensitivity of hNa V 1.4 F1705I could explain myotonia, but it is not clear what triggers cold-induced symptoms or why these patients are free from symptoms at normal body temperature. Our study indicates hastening of I Na decay of hNa V 1.4 F1705I at normal body temperature eliminates current changes that are associated with the myotonic phenotype. At 37uC, the I Na decay of hNa V 1.4 F1705I (at 230 mV) is ,5 times faster than at RT and is nearly identical to the decay rate of wild type hNa V 1.4 current at 37uC (Figures 2C  and 2D). Additionally, at 37uC, V 1/2 of steady-state inactivation of hNa V 1.4 F1705I significantly shifted ,10 mV in hyperpolarizing direction compared to RT, and is comparable to V 1/2 of wild type channel at RT ( Figure 2E). The temperature dependent shift we observed in the human mutant channel is more than the previously reported ,3 mV temperature dependent shift in steady-state inactivation of native rat skeletal muscle Na channels [39]. Rescue of I Na decay and steady-state inactivation at 37uC could explain why the patients with the F1705I mutation are free from myotonia at normal body temperature. Slowing of deactivation of hNa V 1.4 may cause sustained skeletal muscle contraction by delaying repolarization thereby prolonging action potential duration [34]. Differential temperature-induced changes in mutant channel gating; particularly slowing of deactivation resulting in persistent membrane depolarization has been proposed for some hNa V 1.4 mutations causing myotonia [40]; however, the F1705I mutation does not alter deactivation [33]. Additionally, it has been suggested that at normal temperatures more wild type channels are activated compared to mutant channels and with lowering of temperature mutant channels dominate membrane excitability [41]. Alternatively changes in temperature may be associated with changes in intracellular [Ca 2+ ] [42] that exaggerate the inactivation gating defects of the mutant channel. We speculate that temperature sensitive changes in intracellular Ca 2+ could alter cold aggravated myotonia in F1705I mutant muscle.
The biophysical effects of hNa V 1.4 F1705I and the orthologous rat mutation rNa V 1.4 F1698I are consistent with a key role for the CT in inactivation gating. Residues in and around the EFL motif [21,22] of the CT of hNa V 1.5 and Na V 1.1 regulate the Ca 2+ sensitivity of channel gating [14,17,18,23,24,25], which may occur at low levels of free [Ca 2+ ] in the cell [43]. In addition to the EFL, sites in the III-IV linker are involved in Ca 2+ /CaM mediated regulation of gating [28,29]. We exploited the differences in the EFL sequences and Ca 2+ sensitivity of inactivation gating of the human and rat orthologous Na V 1.4 channels to better understand the mechanisms of Ca 2+ modulation of wild type and mutant channels. Inactivation gating of hNa V 1.4 F1705I is sensitive to changes in intracellular [Ca 2+ ], in contrast wild type hNa V 1.4 exhibits no such sensitivity to Ca 2+ . The inverse is true for the wild type and mutant rat isoforms of Na V 1.4. Mutating the nonconserved amino acids of the hNa V 1.4 EFL region to match that of the rat sequence (G1613S and A1636D), abolishes the Ca 2+ sensitivity of hNa V 1.4 F1705I+GS/AD mimicking that of rNa V 1.4 F1698I . Similarly, mutating the non-conserved amino acids of the wild type hNa V 1.4 EFL region to match that of the rat sequence (G1613S and A1636D) restores the Ca 2+ sensitivity of the channel mimicking that of wild type rat Na V 1.4 ( Figures 6G,  6H, Table 1). To our knowledge, this is the first report of a mutation in the CT remote from the EFL region that abolishes Ca 2+ sensitivity of gating of Na V 1.4 channels; however, it is not clear that the rNa V 1.4 F1698I or hNa V 1.4 F1705I has compromised channel gating through a long range conformational effect on the EFL motif. Notably, both rat and human myotonic mutant channels dramatically shift the voltage dependence of inactivation gating in the depolarizing direction. A similar depolarizing shift in inactivation due to a mutation in the CT hNa V 1.4 (Q1633E) has been reported, implicating the EFL region of the CT in inactivation gating [9]. Thus it may be that these mutants introduce a significant local structural change in the predicted H5 which in turn influences the neighboring H4 and alters helical interactions in the EFL leading to a disruption of Ca 2+ sensing. Hydrophobic helical interactions in the cardiac Na V 1.5 EFL, particularly H1-H4, appear to stabilize the structure of the proximal CT of the channel which is postulated to be a prerequisite for durable inactivation [44]. Although most of the mutations at the hydrophobic interfaces of the EFL helices shift inactivation in the hyperpolarizing direction, there are important exceptions to this generalization [44].
The mechanism of the Ca 2+ regulation of gating by residues in the CT is not fully understood and the role of direct Ca 2+ binding to this region is debated. In fact the crystal structure of a ternary complex of the CT-hNa V 1.5, CaM and FHF fails to show Ca 2+ in the EFL region [27,30]. The structure needs to be viewed in the context of functional studies of the intact channel where mutations of the EFL residues consistently alter the Ca 2+ sensitivity of hNa V 1.5 gating [14,17,18,24,45]. This does not prove that Ca 2+ is binding to this region; an alternative is that the mutation(s) produce an allosteric change that alters Ca 2+ sensitivity of gating.
It is interesting that this myotonic mutation in H5 does not affect CaM regulation of the channel. This finding is consistent with previous observations [15] that CaM is an integral part of Na V 1.4 and it remains tethered to the mutant channels in all conformations. An alternative explanation is that the IQcontaining H6 interacts with the EFL [24,45] bringing the H5 into physical proximity of the EFL. Thus mutations such as rNa V 1.4 F1698I or hNa V 1.4 F1705I in H5 could alter the conformation and disrupt Ca 2+ -sensing but this would have to occur without altering the interaction between CaM and the IQ as myotonic mutant channels of both the species are normally regulated by over expression of CaM (Table 1).
It is clear from our study that intracellular Ca 2+ has the capacity to regulate some skeletal muscle Na V 1.4 isoforms in a manner similar to the cardiac isoform Na V 1.5 [14,17,18,24,32], provided the Na V 1.4 channel isoforms have proper residues in the EFL sequence. For example, channels with the rNa V 1.4 EFL ( Figure 6A) or engineered mutant human channels with the orthologous rat residues in the EFL, hNa V 1.4 GS/AD , are regulated by [Ca 2+ ] i . In contrast, channels with the native human sequence (hNa V 1.4) or mutated residues in the EFL (rNa V 1.4 SG/DA ) are insensitive to [Ca 2+ ] i . This study also demonstrates that detailed regulation of inactivation of variants of Na V 1. 4 channels by CaM over expression is determined by the composition of the EFL and the ability of intracellular Ca 2+ to regulate gating (Figure 7). Irrespective of Na V 1.4 channel isoform, CaM mediated regulation of Na V 1.4 depends on the EFL residues and [Ca 2+ ] i . The voltage dependence of inactivation gating of the skeletal muscle Na channel variants is either Ca 2+ sensitive (rat Na V 1. 4, hNa V 1.4 GA/ DS ) or insensitive (human Na V 1. 4, rNa V 1.4 DS/GA ). The Ca 2+ sensitivity is in part determined by the amino acid sequence in the EFL. This sequence divergence is associated with differences in the baseline voltage dependence of inactivation as well as the response to CaM over expression (Figure 7).

Conclusion
The cold aggravated myotonia mutation, hNa V 1.4 F1705I in the CT of the skeletal muscle channel remote from the EFL region produces temperature-dependent slowing of current decay and significant destabilization of inactivation, and is associated with a disruption or alteration of Ca 2+ regulation. Alteration of [Ca 2+ ] i sensitivity at low temperature could potentiate myotonia symptoms. These changes result in greater Na current availability at depolarized voltages, and thus prolongation of the cellular action potential which is likely to be the proximate cause of myotonia. Moreover the data suggests a mechanism by which drugs that stabilize Na current inactivation may be useful in controlling muscle symptoms. This disease causing mutation and isoform specific amino acid variation in the CT EFL provide important insights into the differences in Ca 2+ regulation of Na V 1 channels.