A skeletal muscle L-type Ca2+ channel with a mutation in the selectivity filter (CaV1.1 E1014K) conducts K+

A glutamate-to-lysine substitution at position 1014 within the selectivity filter of the skeletal muscle L-type Ca2+ channel (CaV1.1) abolishes Ca2+ flux through the channel pore. Mice engineered to exclusively express the mutant channel display accelerated muscle fatigue, changes in muscle composition, and altered metabolism relative to wildtype littermates. By contrast, mice expressing another mutant CaV1.1 channel that is impermeable to Ca2+ (CaV1.1 N617D) have shown no detectable phenotypic differences from wildtype mice to date. The major biophysical difference between the CaV1.1 E1014K and CaV1.1 N617D mutants elucidated thus far is that the former channel conducts robust Na+ and Cs+ currents in patch-clamp experiments, but neither of these monovalent conductances seems to be of relevance in vivo. Thus, the basis for the different phenotypes of these mutants has remained enigmatic. We now show that CaV1.1 E1014K readily conducts 1,4-dihydropyridine-sensitive K+ currents at depolarizing test potentials, whereas CaV1.1 N617D does not. Our observations, coupled with a large body of work by others regarding the role of K+ accumulation in muscle fatigue, raise the possibility that the introduction of an additional K+ flux from the myoplasm into the transverse-tubule lumen accelerates the onset of fatigue and precipitates the metabolic changes observed in CaV1.1 E1014K muscle. These results, highlighting an unexpected consequence of a channel mutation, may help define the complex mechanisms underlying skeletal muscle fatigue and related dysfunctions.

During excitation-contraction (EC) 2 coupling in skeletal muscle, the L-type Ca 2ϩ channel (Ca V 1.1 or 1,4-dihydropyridine receptor) activates Ca 2ϩ release from the sarcoplasmic reticulum via the type 1 ryanodine receptor in response to depolarization of the plasma membrane (1)(2)(3). Because EC cou-pling is fast and does not appear to depend upon a soluble second messenger (e.g. Ca 2ϩ ), there is general agreement that there is direct or indirect conformational coupling between these two channels within a larger macromolecular signaling complex (3)(4)(5). In addition to its primary function as an EC coupling voltage sensor, Ca V 1.1 also conducts L-type Ca 2ϩ current (3,6).
To address the importance of Ca 2ϩ flux via Ca V 1.1 for greater muscle function, two distinct mouse lines have been engineered. Both of these strains carried single amino acid substitutions that rendered the channel impermeable to Ca 2ϩ while sparing EC coupling. In one model, Ca V 1.1 had a targeted mutation within the selectively filter (E1014K) (7) known to eliminate nearly all divalent flux (5, 8 -10). Although the amplitude of myoplasmic Ca 2ϩ release evoked by low-frequency stimulation (LFS; 1 Hz) was virtually identical in flexor digitorum brevis (FDB) fibers obtained from wildtype and homozygous Ca V 1.1 E1014K mice, Ca V 1.1 E1014K FDB fibers displayed a pronounced fatigue phenotype because the amplitudes of successive Ca 2ϩ transients decayed more rapidly than in wildtype fibers during high-frequency stimulation (HFS; 50 or 100 Hz) (7). In addition, homozygous expression of the Ca V 1.1 E1014K channel caused a gain of glycolytic type IIB fibers at the expense of type IIX fibers in extensor digitorum longus and type IIX and type I fibers in soleus muscles (7). Ca V 1.1 E1014K mice were later reported to develop multiple metabolic deficiencies leading to increased fat mass and overall body weight (11).
A mouse model based on the non-conducting zebrafish ␣ 1S -b isoform has also been generated (12). These mice expressed a Ca V 1.1 channel having an aspartate for asparagine swap adjacent to the selectivity filter at position 617 (13,14). In stark contrast to Ca V 1.1 E1014K mice, Ca V 1.1 N617D mice had no detectable phenotypic differences from wildtype littermates in a variety of assays. Notably, the authors found no significant effects on body weight, fertility, muscle mass/composition, EC coupling, SR Ca 2ϩ store content, twitch or tetanic force, muscle fatigue following HFS, locomotor function or the expression levels of Ca V 1.1, and other mediators of Ca 2ϩ signaling in skeletal muscle (e.g. RyR1, Orai1, STIM1, SERCA1, CSQ1, CSQ2, etc.).
Although neither Ca V 1.1 E1014K nor Ca V 1.1 N617D conduct appreciable L-type Ca 2ϩ current in either cultured or acutely dissociated muscle cells, the former channel is known to conduct Na ϩ and Cs ϩ currents. Reduction of external Ca 2ϩ in patch-clamp experiments has revealed inward Ca V 1.1 E1014Kmediated Na ϩ currents (10), whereas the ability of Ca V 1.1 E1014K to conduct large-amplitude outward Cs ϩ currents was established during the initial characterization of the mutant (5). The selectivity of Ca V 1.1 N617D for monovalent cations has not yet been defined, but significant outward currents were not observed in experiments performed with 100 -145 mM Cs ϩ present in the patch pipette (12)(13)(14).
The apparent impermeability of Ca V 1.1 N617D to Cs ϩ suggests that the altered selectivity of Ca V 1.1 E1014K underlies the phenotypic differences between Ca V 1.1 E1014K and Ca V 1.1 N617D mice. Obviously, the Cs ϩ permeability of Ca V 1.1 E1014K is inconsequential with regard to phenotype because Cs ϩ in not present in significant quantities in skeletal muscle fibers in vivo. Likewise, Na ϩ flux via Ca V 1.1 E1014K is unlikely to be the basis for these differences because the channel opens much too slowly to conduct Na ϩ during the upstroke of a single action potential (7). However, the ability of Ca V 1.1 E1014K to conduct the other physiologically relevant monovalent cation, K ϩ , has not been determined.
In this work, we expressed YFP-tagged Ca V 1.1 E1014K and Ca V 1.1 N617D channels in tsA-201 cells to demonstrate that Ca V 1.1 E1014K functions as a non-inactivating K ϩ channel, whereas Ca V 1.1 N617D does not. We postulate, based a very large body of earlier work regarding fatigue (reviewed in detail in Ref. 15), that the conversion of Ca V 1.1 to a K ϩ channel within the transverse tubules (triad junctions) of homozygous Ca V 1.1 E1014K muscle results in enhanced K ϩ accumulation leading to accelerated fatigue during periods of high activity. Because changing the selectively of Ca V 1.1 via the E1014K mutation has not been investigated previously as the mechanism of the accelerated fatigue phenotype, our results may help clarify the impact of L-type Ca 2ϩ flux into skeletal muscle.

Successful expression of YFP-Ca V 1.1 E1014K in tsA-201 cells
We first examined whether the Ca V 1.1 E1014K mutant could be functionally expressed in tsA-201 cells with similar biophysical properties as previously reported for Ca V 1.1 E1014K expressed in dysgenic myotubes. For this purpose, we constructed a YFP-Ca V 1.1 E1014K fusion construct similar to the untagged Ca V 1.1 E1014K channel utilized by Dirksen and Beam (5). tsA-201 cells were then transfected with YFP-Ca V 1.1 E1014K or YFP-Ca V 1.1 and ␤ 1a , ␣ 2 ␦-1, and Stac3 auxiliary channel subunits (16); positively transfected cells were identified by YFP fluorescence (Fig. 1A, inset). With 2 mM Ca 2ϩ /150 TEA ϩ in the bath (i.e. TEA-Tyrode's solution; see "Experimental procedures") and ϳ160 mM Cs ϩ in the pipette solution, cells expressing YFP-Ca V 1.1 E1014K yielded no inward Ca 2ϩ or Na ϩ current but supported sizable outward Cs ϩ currents (I dens ϭ ϩ83.4 Ϯ 11.6 pA/pF at ϩ80 mV, n ϭ 18; Fig. 1A) with an current-voltage (I-V) relationship similar to that reported earlier for untagged Ca V 1.1 E1014K expressed in dysgenic myotubes ( Fig. 1C) (5,8). As expected, the cells expressing YFP-Ca V 1.1 displayed typical inward L-type Ca 2ϩ currents peaking near ϩ30 mV (I dens ϭ Ϫ4.8 Ϯ 1.0 pA/pF, n ϭ 12; Fig. 1B). Importantly, the L-type current mediated by YFP-Ca V 1.1 did  Fig. 3A shows a family of currents recorded from a tsA-201 cell expressing YFP-Ca V 1.1 E1014K with TEA-Tyrode's solution in the bath and ϳ150 mM K ϩ in the pipette. No inward L-type Ca 2ϩ or Na ϩ currents were detectable, but non-inactivating outward currents were evident at test potentials greater than 0 mV (I dens ϭ 92.9 Ϯ 14.6 pA/pF at ϩ80 mV; n ϭ 16; Fig.  3, A and D). By contrast, YFP-Ca V 1.1 N617D produced virtually no outward current above the average tsA-201 background (I dens ϭ 3.3 Ϯ 0.5 pA/pF; n ϭ 9; p Ͻ 0.001, unpaired t test versus YFP-Ca V 1.1 E1014K; Fig. 3, B and D). YFP-Ca V 1.1 N617D I-V data were collected only from cells in which channel expression was confirmed by the presence of charge movement (Q max ϭ 12.8 Ϯ 1.9 nC/F, n ϭ 9). Because YFP-Ca V 1.1 N617D actually displayed slightly larger charge movement than YFP-Ca V 1.1 or YFP-Ca V 1.1 E1014K (p Ͼ 0.05, one way-ANOVA; Fig. S1), the lack of K ϩ current in cells expressing Ca V 1.1 N617D could not be attributed to fewer channels present in the membrane. Under these conditions, YFP-Ca V 1.1 produced L-type currents similar in amplitude and voltage dependence to the experiments shown in Fig. 1 in which Cs ϩ was used as the predominant intracellular monovalent cation (Fig. 3, C and D). The reversal potential was shifted ϳ10 in the hyperpolarizing direction with K ϩ in the pipette (V rev ϭ 70.1 Ϯ 2.9 mV; p Ͻ 0.05), an effect likely caused by an incomplete block of endogenous K ϩ channels.

K ؉ currents conducted by YFP-Ca V 1.1 E1014K are reduced by nifedipine
The effects of DHP agonists and antagonists on Ca V 1.1 E1014K are complex. For example, the agonist ϮBay K 8644 amplifies inward Na ϩ currents (10) but causes potentiated outward Cs ϩ currents to rectify (10,17). To determine whether the outward K ϩ current is sensitive to inhibition by DHP antago-

Ca V 1.1 E1014K conducts K ؉
nists, YFP-Ca V 1.1 E1014K-expressing tsA-201 cells were exposed acutely to nifedipine. Fig. 4A shows K ϩ currents recorded at ϩ80 mV from the same YFP-Ca V 1.1 E1014K-expressing cell before and during sequential applications of 500 nM and 10 M nifedipine; the time course of inhibition in this particular experiment is depicted in Fig. 4B. These concentrations of nifedipine reduced the total K ϩ current by 27.0 Ϯ 3.6% (n ϭ 6; p Ͻ 0.05, paired t test versus control) and by 78.8 Ϯ 3.3% (n ϭ 6; p Ͻ 0.05, paired t test), respectively (Fig. 4C). The degrees of inhibition by both 500 nM and 10 M nifedipine were somewhat inconsistent with earlier reports indicating that nifedipine inhibits L-type currents in myotubes with an IC 50 near 500 nM (8,18). Upon reflection, it became apparent that the previous studies employed depolarizing steps from the steady holding potential immediately prior to the test step to inactivate the endogenous Na ϩ and T-type Ca 2ϩ channels of myotubes (steps ranging from Ϫ50 to Ϫ20 mV for durations varying between 50 and 1000 ms). Because the tsA-201 cells do not express Na ϩ or T-type channels in any abundance, we had been stepping directly from the steady holding potential of Ϫ80 mV to test potential of ϩ80 mV without taking into account that DHP affinity is directly dependent on holding potential (Ref. 19; for a review please see Ref. 20). Therefore, we performed an additional experiment in which we stepped to Ϫ50 mV for 750 ms prior to the test potential. As shown in Fig. 4D, the degree of inhibition following the step to Ϫ50 mV was in much better agreement with the aforementioned studies (47.3 Ϯ 2.7 and 90.2 Ϯ 1.1% inhibition with 500 nM nifedipine and 10 M nifedipine, respectively; both n ϭ 6).
Thus, the bulk of the K ϩ current arose primarily from the expressed Ca V 1.1 E1014K channel rather than via endogenous K ϩ channels.

Bi-ionic current recordings from YFP-Ca V 1.1 E1014K
Because Ca V 1.1 E1014K channels conduct Na ϩ (10) as well as K ϩ , we recorded hybrid Na ϩ and K ϩ currents with regular Tyrode's solution in the bath. In these experiments, we observed both inward Na ϩ and outward K ϩ current with a biionic reversal potential ϳϩ20 mV (n ϭ 10; Fig. 5A). No inward nor appreciable outward currents were observed in seven recordings made from cells expressing Ca V 1.1 N617D under identical conditions (Fig. 5B). The bionic current-voltage relationships for tsA-201 cells expressing either YFP-Ca V 1.1 E1014K or YFP-Ca V 1.1 N617D are shown in Fig. 5C.

Discussion
In this study, we used the newly developed tsA-201 cell heterologous expression system to demonstrate that the Ca V 1.1  . Bi-ionic current families shown were evoked by 100-ms steps from Ϫ80 mV to Ϫ30-, Ϫ10-, ϩ10-, ϩ30-, ϩ50-, ϩ70-, and ϩ90-mV increments. The blue Na ϩ and red K ϩ denote the inward and outward charge carriers, respectively. Average bi-ionic I-V relationships for tsA-201 cells expressing either Ca V 1.1 E1014K (black circles; n ϭ 10) or Ca V 1.1 N617D (gray circles; n ϭ 7) are shown in C.

Ca V 1.1 E1014K conducts K ؉
E1014K mutant channel will readily conduct K ϩ (Fig. 3). Although Ca V 1.1 E1014K is known to conduct Na ϩ current (10), L-type Ca 2ϩ channels are unlikely to be open during the upstroke of a single muscle action potential when the membrane potential favors the influx of Na ϩ (7). The same cannot be said unequivocally for the more depolarized stages of the action potential when the driving force for K ϩ is strong (21,22). Thus, Ca V 1.1 E1014K is likely to function as a K ϩ channel when knocked in to skeletal muscle, particularly during repetitive activity or prolonged depolarization when the channel enters high P o gating modes (17,23).
We have also found that the outward K ϩ current mediated by Ca V 1.1 E1014K has a similar sensitivity to the DHP antagonist nifedipine (Fig. 4) as has been reported earlier for native and heterologously expressed wildtype Ca V 1.1 channels in cultured myotubes (8,18). The sensitivity of Ca V 1.1 E1014K to nifedipine in our voltage-clamp experiments contrasts with in vitro work showing that the mutation reduces the affinity of the channel for radiolabeled PN200-100 (isradipine) in membrane preparations obtained from tsA-201 cells expressing either wildtype or mutant Ca V 1.1 ␣ 1S with only ␤ 1a and ␣ 2 ␦-1 subunits (24). The results of our experiments indicate that E1014K mutation does not greatly affect DHP sensitivity of Ca V 1.1 in a cellular context, as suggested by the binding assays of Peterson and Catterall (24).
To investigate the impact of the E1014K mutation on muscle activity, Lee et al. (7) loaded Ca V 1.1 E1014K FDB fibers with Mag-Fluo4 AM dye and then recorded myoplasmic Ca 2ϩ transients elicited by field stimulation. They found that amplitudes and durations of the Ca 2ϩ transients remained constant for both wildtype and Ca V 1.1 E1014K fibers in response to 1 Hz LFS, indicating that the basic mechanism of EC coupling was not likely affected by the mutation. During 50 or 100 Hz HFS, the amplitudes of successive Ca 2ϩ transients in wildtype fibers decreased steadily, whereas the amplitudes of Ca 2ϩ transients in Ca V 1.1 E1014K fibers decayed more rapidly (see Fig. 2, A-C, of Ref. 7). In this regard, a role for K ϩ accumulation within the transverse-tubule lumen and other extracellular compartments has been described as a major contributor to action potential/EC coupling failure and subsequent declines in force during some HFS protocols (15,(25)(26)(27)(28)(29)(30). During the repolarization phase of the skeletal muscle action potential, K ϩ exits the myoplasm via delayed rectifier K ϩ channels to restricted extracellular compartments before being returned actively to the fiber by the Na ϩ /K ϩ -ATPase (31) or passively via inward rectifier K ϩ channels when the membrane potential becomes more negative than E K (32)(33)(34)(35). During LFS, these clearance mechanisms are sufficient to maintain the K ϩ gradient (31). However, transport by the pump and inward rectifier channels cannot keep pace with the K ϩ efflux via delayed rectifier channels during prolonged depolarization or HFS protocols similar to those employed by Lee et al. (15,30,(33)(34)(35)(36)(37)(38)(39). Even with the assistance of Cl Ϫ flux via ClC-1 channels in stabilizing the resting potential, the result is net K ϩ accumulation sufficient to elevate the local membrane potential to a point at which Na ϩ channels inactivate (30,40). Failure of EC coupling follows because of: 1) the inability of the fiber to propagate action potentials (30,41) and 2) inactivation of Ca V 1.1 itself (42).
Based on our observation that Ca V 1.1 E1014K functions as a triad junction-targeted K ϩ channel (Figs. 3 and 5), we propose a relatively simple model that may explain why failure of EC coupling ensues more rapidly in response to repetitive HFS in Ca V 1.1 E1014K muscle than in wildtype or Ca V 1.1 N617D muscle in Fig. 6. As described above, endogenous delayed rectifier K ϩ channels provide an avenue for K ϩ efflux into the lumen during HFS or prolonged depolarization in wildtype muscle (Fig. 6A) (34,35,43). The introduction of a substantial K ϩ conductance via every single Ca V 1.1 channel in Ca V 1.1 E1014K muscle would accelerate the elevation of the local membrane potential, resulting in action potential and EC coupling failure leading to more rapid declines in force (Fig. 6B). In particular, the probability of EC coupling failure is high because of the high density of Ca V 1.1 channels in the triad junction subcompartment (44,45); the abundance of delayed rectifier channels at triad junctions has not been determined, but we have chosen arbitrarily to include these channels at low density relative to Ca V 1.1 in our model (as opposed to omitting them altogether). Because Ca V 1.1 N617D, like wildtype Ca V 1.1, does not conduct appreciable K ϩ at physiological potentials (Fig. 3, B-D), K ϩ Figure 6. Simplified model illustrating how additional K ؉ flux via Ca V 1.1 E1014K from the myoplasm into the transverse-tubule lumen (triad junctions) may exacerbate muscle fatigue following intense activity. A, left panel, a cartoon depicting K ϩ (green circles) levels in the myoplasm (gray compartments) and transverse-tubule lumen (colorless area) of wildtype muscle at rest. In all panels, wildtype and mutant Ca V 1.1 channels are represented by blue cylinders with black Xs indicating the inability to conduct K ϩ . Right panel, K ϩ accumulation in the transverse-tubules of wildtype muscle following HFS (illustrated in red, center). Here, the delayed rectifier K ϩ channels (red cylinders) of wildtype muscle provide the primary route for K ϩ flux into the confined space of the transverse-tubules leading to muscle fatigue after HFS (15). B, left panel, Ca V 1.1 E1014K muscle at rest. B, right panel, enhanced K ϩ accumulation in the transverse-tubules of Ca V 1.1 E1014K muscle following HFS. The fatigue process is exacerbated by the addition of another K ϩ conductance carried by Ca V 1.1 E1014K. C, left panel, Ca V 1.1 N617D muscle at rest. C, right panel, during HFS, K ϩ accumulates in the transverse-tubules of Ca V 1.1 N617D muscle at a rate similar to wildtype muscle (A, right panel) because Ca V 1.1 N617D does not conduct K ϩ . For clarity, K ϩ removal routes (e.g. inward rectifier K ϩ channels, Na ϩ /K ϩ ATPase pump, diffusion, etc.) and Cl Ϫ conductances have been omitted. Ca V 1.1 E1014K conducts K ؉ accumulation proceeds at a rate similar to wildtype muscle (Fig. 6C).
The most direct way to test the idea that K ϩ flux via Ca V 1.1 E1014K accentuates K ϩ accumulation in extracellular compartments would be to assess the reversal potentials of the inward rectifier currents in wildtype, Ca V 1.1 E1014K, and Ca V 1.1 N617D fibers before and immediately following HFS or long depolarizations (33)(34)(35). Specifically, a higher conductance and a more profound depolarizing shift in the reversal potential for the inward rectifier current in Ca V 1.1 E1014K fibers relative to wildtype and Ca V 1.1 N617D fibers would be evident if elevated extracellular K ϩ levels underlie the accelerated decay of Ca 2ϩ transient amplitudes during HFS. Because our data indicate that nifedipine inhibits K ϩ flux via Ca V 1.1 E1014K (Fig. 4), the increase in the conductance and the shift in the reversal potential of the inward rectifier current would be reduced or eliminated by application of a DHP antagonist.
The assertion of Lee et al. (7) that Na ϩ influx is via Ca V 1.1 E1014K is not likely during the upstroke of a single action potential is almost certainly correct, but it is important to note that our results do not completely exclude a role for Na ϩ flux via Ca V 1.1 E1014K during the later stages of the action potential. However, it would seem that the depolarizing influence of a Na ϩ tail current in Ca V 1.1 E1014K muscle would be similar to that of Ca 2ϩ tail current in wildtype muscle (Fig. 3C). Likewise, the Goldman-Hodgkin-Katz relationship implies that an additional Na ϩ flux into the relatively large volume of the myoplasm would not have nearly the same weight on resting membrane potential as K ϩ efflux into the confined space of the transverse tubules, which only account for Ͻ 0.5% of the total fiber volume (43,46,47). Related to this point, currents recorded in nearly equimolar Na ϩ external and K ϩ internal displayed a bi-ionic reversal potential of ϳ20 mV, which was depolarized ϳ15 mV relative to potentials where outward K ϩ currents became visible in the absence of external Na ϩ (compare Fig. 5C with Fig.  3D). Because the bi-ionic reversal potential is the voltage in which the fluxes of Na ϩ and K ϩ are in equilibrium, this parameter does not provide information regarding the absolute flux of either individual ion, other than that the fluxes are equal and opposite at that potential (47). In regard to the model presented in Fig. 6, a very negative Nernst potential (ϳϪ98 mV in skeletal muscle) (47) dictates that K ϩ will exit the myoplasm in vivo at potentials hyperpolarized to the bi-ionic reversal potential if a K ϩ ion occupies the selectivity filter of the open channel.
If Ca V 1.1 E1014K supports K ϩ efflux during repetitive activity in vivo, the implication is that there is physiological Ca 2ϩ flux via the wildtype Ca V 1.1 channel. Based on the observations of Dayal et al. (12) indicating that the Ca V 1.1 N617D substitution has no detectable physiological consequences, the conclusion that L-type Ca 2ϩ flux is vestigial in muscle of mice and, by extrapolation, other mammals, appears correct (but see Ref. 22). Although our findings support the idea that the N617D mutation is without effect because it simply eliminates an ionic flux that is of very little or no consequence, the fatigue phenotype of Ca V 1.1 E1014K muscle can be linked reasonably to the additional K ϩ flux that occurs as a result of the altered selectivity of the latter channel. What is not so clear is how this additional K ϩ conductance could cause fiber atrophy, fiber-type switching, altered fatty acid metabolism, decreased energy expenditure, and gain of fat mass in mice over the course of weeks (7,11). In view of the apparent relationship between aberrant membrane excitability in skeletal muscle and metabolic dysfunction, the need for further investigation is obvious and pressing; our findings provide a necessary first step toward making this connection.

Molecular biology
YFP fused-Ca V 1.1 E1014K and -Ca V 1.1 N617D were derived from the plasmid YFP-Ca V 1.1 (48). The generation of YFP-Ca V 1.1 E1014K was undertaken by Genscript, Inc. Four point mutations (ttc gag to ttt aaa) spanning rabbit Ca V 1.1 (GenBank TM accession number X05921) bp 3037-3042 were introduced into YFP-Ca V 1.1 through two rounds of PCR mutagenesis using the 4125-bp template excised from YFP-Cav1.1 with EcoRI-PmlI restriction enzymes. The sequence of the forward primers were 5Ј-cggactcagatctcgagctcaagcttcgaattctgcagtcgacatg-3Ј and 5Ј-ttcacggtgtccacctttaaaggatggccccagctgctgtacaggg-3Ј. The sequence of the reverse primers were 5Јcagctggggccatcctttaaaggtggacaccgtgaagagcga-3Ј and 5Ј-aagtagtagtaggcgaagttggtgccacacgtgtactcct-3Ј. These primers were used to generate a new 4125-bp fragment containing the four mutations that was then subcloned back into the original vector using EcoRI-PmlI restriction sites. Like the generation of the original untagged Ca V 1.1 E1014K clone in pCAC6 generated by Dirksen and Beam (SkEIIIK) (5), our strategy not only swapped in a lysine codon (aaa) in place of the glutamate codon (gag) at position 1014 but also modified the preceding phenylalanine codon (ttt for ttc) at position 1013 to introduce a DraI restriction site (ttt aaa). The completed construct was verified first by XhoI restriction digest and sequencing at Genscript and later by the presence of the introduced DraI restriction site upon arrival at the University of Colorado.
For YFP-Ca V 1.1 N617D, a single a-to-g point mutation at bp 1849 of rabbit Ca V 1.1 was introduced into YFP-Ca V 1.1 using the QuikChange II XL site-directed mutagenesis kit (Agilent Technologies). The sequence of the forward primer was 5Јacgggtgaggactgggactccgtgatgtacaac-3Ј. The sequence of the reverse primer was 5Ј-gttgtacatcacggagtcccagtcctcacccgt-3Ј. The completed construct was verified by sequencing at the DNA Sequencing Core of the Barbara Davis Center at the University of Colorado. YFP-Ca V 1.1 and YFP-Ca V 1.1 N617D expression plasmids were both gifts from Dr. K. G. Beam (University of Colorado School of Medicine).

Ionic current recordings
Pipettes were fabricated from borosilicate glass and had resistances of 2.0 -4.0 M⍀ when filled with internal solution. For Cs ϩ current recordings, the internal solution consisted of 140 mM cesium aspartate, 10 mM Cs 2 -EGTA, 5 mM MgCl 2 , and 10 mM HEPES, pH 7.4, with CsOH. For K ϩ current recordings, the internal solution consisted of 130 mM potassium asparatate, 20 mM KCl, 1 mM MgCl 2 , 10 mM EGTA, and 10 mM HEPES, pH 7.4, with KOH. To record Na ϩ /K ϩ bi-ionic current-voltage relationships, the external solution contained 137 mM NaCl, 4 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 10 mM HEPES, 10 mM glucose, and 1 mM 4-aminopyridine, pH 7.4, with NaOH. Otherwise, the external solution contained 145 mM TEA-Cl, 4 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 10 mM HEPES, 10 mM glucose, and 1 mM 4-aminopyridine, pH 7.4, with TEA-OH. The latter two solutions are referred to as Tyrode's solution and TEA-Tyrode's solution, respectively, in the text of "Results." All ionic current recordings were corrected for linear cell capacitance and leakage currents using an on-line ϪP/4 subtraction protocol. K ϩ or Cs ϩ currents were filtered at 2 kHz and digitized at 5 kHz. Cell membrane capacitance (C m ) was determined by integration of a transient from Ϫ80 to Ϫ70 mV using Clampex 8.0 or 10.3 (Molecular Devices) and was used to normalize current amplitudes (pA/pF). The time constant for decay of the whole-cell capacity transient ( m ) was reduced as much as possible using the analog compensation circuit of an Axon 200B amplifier (Molecular Devices). The average values of C m , m , and access resistance (R a ) for all recordings were 42.3 Ϯ 3.1 pF, 374 Ϯ 42 s, and 9.3 Ϯ 0.6 M⍀, respectively (n ϭ 98 cells). Where applicable, L-type Ca 2ϩ I-V relationships were fitted using the following equation: where I is the current for the test potential V, V rev is the reversal potential, G max is the maximum channel conductance, V1 ⁄ 2 is the half-maximal activation potential, and k G is the slope factor. Nifedipine (TCI Chemicals) was dissolved in 100% EtOH to make a 10 mM stock solution and then diluted to 10 M and 500 nM in TEA-Tyrode's solution just prior to experiments. During experiments, nifedipine working solutions were applied through a manually operated, gravity-driven global perfusion system.

Charge movement recordings
To record charge movements from YFP-Ca V 1.1 and YFP-Ca V 1.1 E1014K, the Cs ϩ -based internal recording solution was used (see "Ionic current recordings"). For YFP-Ca V 1.1 N617D, charge movements were recorded with the same internal solution used to record K ϩ currents. In all three cases, the bath solution contained 145 mM TEA-Cl, 4 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 0.1 mM LaCl 3 , 0.5 mM CdCl 2 , 10 mM HEPES, 10 mM glucose, and 1 mM 4-aminopyridine, pH 7.4, with TEA-OH. Linear components of leak and capacitive current were corrected with ϪP/4 online subtraction protocols. Output filtering was at 5-10 kHz, and digitization was at 25 kHz. For analysis, the ON component of the charge transient (Q ON ) was normalized to C m and plotted as a function of test potential (V), and the resultant Q-V relationships were fitted according to the following equation, where Q max is the maximal Q ON , V Q is the potential causing movement of half the maximal charge, and k Q is a slope parameter. All experiments were performed at room temperature (ϳ23°C).

Confocal imaging
Images of live tsA-201 cells expressing either YFP-Ca V 1.1 E1014K or YFP-Ca V 1.1 N617D were acquired as described previously (50).

Analysis
The figures were made using the software program Sigma-Plot (version 11.0, Systat Software, Inc.). All data are presented as means Ϯ S.E. All statistical comparisons were by unpaired, two-tailed t test, unless otherwise noted. p Ͻ 0.05 was considered significant.
Author contributions-D. B. and R. A. B. designed the research, performed the research, analyzed the data, and wrote the paper. K. D. performed the research. All authors read and approved the final manuscript.