Initiation of skeletal muscle contraction is triggered by rapid activation of RYR1 channels in response to sarcolemmal depolarization. RYR1 is intracellular and has no voltage-sensing structures, but it is coupled with the voltage-sensing apparatus of CaV1.1 channels to inherit voltage sensitivity. Using an opto-electrophysiological approach, we resolved the excitation-driven molecular events controlling both CaV1.1 and RYR1 activations, reported as fluorescence changes. We discovered that each of the four human CaV1.1 voltage-sensing domains (VSDs) exhibits unique biophysical properties: VSD-I time-dependent properties were similar to ionic current activation kinetics, suggesting a critical role of this voltage sensor in CaV1.1 activation; VSD-II, VSD-III, and VSD-IV displayed faster activation, compatible with kinetics of sarcoplasmic reticulum Ca2+ release. The prominent role of VSD-I in governing CaV1.1 activation was also confirmed using a naturally occurring, charge-neutralizing mutation in VSD-I (R174W). This mutation abolished CaV1.1 current at physiological membrane potentials by impairing VSD-I activation without affecting the other VSDs. Using a structurally relevant allosteric model of CaV activation, which accounted for both time- and voltage-dependent properties of CaV1.1, to predict VSD-pore coupling energies, we found that VSD-I contributed the most energy (~75 meV or ∼3 kT) toward the stabilization of the open states of the channel, with smaller (VSD-IV) or negligible (VSDs II and III) energetic contribution from the other voltage sensors (<25 meV or ∼1 kT). This study settles the longstanding question of how CaV1.1, a slowly activating channel, can trigger RYR1 rapid activation, and reveals a new mechanism for voltage-dependent activation in ion channels, whereby pore opening of human CaV1.1 channels is primarily driven by the activation of one voltage sensor, a mechanism distinct from that of all other voltage-gated channels.

The contraction of skeletal muscles is initiated by a brief (3–5 ms) action potential that causes a depolarization of muscle fibers (Bannister, 2016; Shishmarev, 2020). This sudden change of the sarcolemma membrane potential is detected by CaV1.1 channel voltage-sensing domains (VSDs), which are compelled to move by the change in electric field. The ensuing structural rearrangements of the VSDs not only open the CaV1.1 pore (allowing Ca2+ to enter the cell) but also propagate to ryanodine receptor 1 (RYR1) channel, located intracellularly in the membrane of the SR. As a result, the RYR1 channel, which cannot directly “sense” an electrical excitation, acquires voltage dependence by physically coupling with a voltage-sensing protein (CaV1.1), such that sarcoplasmic Ca2+ release and muscle contraction become voltage-dependent (depolarization-driven) events (Rios and Brum, 1987; Ríos et al., 1992).

Seminal studies in the early 1970s recognized that an intramembrane “charge movement” was associated with, and necessary for, the skeletal muscle excitation–contraction (EC) coupling process (Schneider and Chandler, 1973; Hernández-Ochoa and Schneider, 2018). This charge movement is indeed the electrical manifestation of the voltage-dependent conformational changes of CaV1.1 voltage-sensing machinery composed of four VSDs (Fig. 1 A, gray trace). In this study, we sought to resolve the independent conformational rearrangements of the four CaV1.1 VSDs to answer a fundamental question in muscle physiology: how does the CaV1.1 channel voltage-sensing machinery govern two different events with very distinct voltage- and time-dependent properties, i.e., the slow opening of CaV1.1 pore and the fast activation of RYR1? In fact, CaV1.1 pore activation takes several tens of milliseconds to reach steady-state open probability (Fig. 1 A, red ionic current trace), while the kinetics of SR Ca2+ release through RYR1 is ∼25-fold faster than pore opening (Fig. 1 B, black trace). We postulated that the four homologous but not identical CaV1.1 VSDs possess different functional properties, enabling CaV1.1 to control both fast and slow molecular events.

Indeed, CaV channels are pseudo-tetramers composed of four concatenated repeats, each encompassing six transmembrane segments (S1–S6; Fig. S1 A): S1–S4 encode for the voltage sensors, while S5 and S6 from each repeat form the central pore (Dolphin, 2016; Hering et al., 2018; Catterall et al., 2020). Thus, a unique amino acid sequence pertains to each voltage sensor (Fig. S1 B), likely underpinning their distinct operation, as previously shown for the close relative CaV1.2 channel (Pantazis et al., 2014; Savalli et al., 2016) and for two NaV isoforms (Cha et al., 1999; Chanda and Bezanilla, 2002; Varga et al., 2015). Until recently, studies of CaV1.1 biophysical properties have been hampered by negligible expression level of this channel in heterologous systems. With the discovery of Stac3-mediated trafficking of CaV1.1 channels (Polster et al., 2015; Flucher and Campiglio, 2019), it is now possible to express these proteins at the plasma membrane both in HEK cells (Polster et al., 2015; Niu et al., 2018) and in Xenopus laevis oocytes (Wu et al., 2018). We used the voltage-clamp fluorometry (VCF) technique (Mannuzzu et al., 1996; Gandhi and Olcese, 2008) to resolve the depolarization-induced structural changes of the four CaV1.1 VSDs in a cellular environment. We found that each VSD of human CaV1.1 channels exhibits distinct biophysical properties. VSD-I time-dependent properties were strikingly similar to ionic current activation kinetics, while VSDs II–IV displayed faster activation, compatible with SR Ca2+-release kinetics, suggesting that VSD-I plays a major role in pore opening. The relevance of VSD-I to CaV1.1 activation was reinforced using a naturally occurring, charge-neutralizing mutation in VSD-I (R174W; Eltit et al., 2012), which ablates CaV1.1 current at physiological membrane potentials by stabilizing VSD-I resting state. Using a structurally relevant allosteric model of CaV activation, which accounted for both time- and voltage-dependent properties of CaV1.1, we quantified the energetic interactions of each VSD with pore opening. Overall, this study revealed a new mechanism for voltage sensing among the voltage-gated superfamily of ion channels, whereby the open state of the human CaV1.1 channel is mainly stabilized by one voltage sensor, leaving the other three VSDs available for other voltage-dependent tasks, such as RYR1 activation.

### Ca2+ release measurement

Enzymatically dissociated flexor digitorum brevis fibers from mice were loaded with the dye OGB-5N (250 µM) and voltage-clamped using the two-electrode voltage-clamp technique. The intracellular solution was (in mM) 75 aspartate, 5 ATP-Na2, 5 phospho-creatine di-Tris, 5 reduced glutathione, 5 MgCl2, 30 EGTA, 15 Ca(OH)2, 20 3-(N-morpholino)propanesulfonic acid, and 4 MOPS, pH 7.4 with KOH. The extracellular solution was (in mM) 150 TEA-Cl, 10 CsCl, 2 CaCl2, 1 MgCl2, and 10 MOPS, pH 7.4. Fluorescence was filtered at 2 kHz.

### Xenopus oocytes isolation

A Xenopus frog is anesthetized by immersion in water containing 0.17% tricaine for ∼30 min. The anesthetized frog is euthanized by central nervous system pithing and laid supine on ice. Ovarian lobes are taken out and placed in sterile OR-2 solution (in mM: 82.5 NaCl, 2.5 KCl, 1 MgC12, and 5 HEPES, pH 7). After shredding the ovarian lobes, ∼5 ml shredded ovaries are incubated with 20 ml of fresh collagenase type-I solution (207 U/ml dissolved in OR-2) in a 50-ml centrifuge tube. The tube is secured horizontally to an orbital shaker and shaken vigorously for 10–20 min to enzymatically and mechanically defolliculate oocytes. After rinsing five or six times with OR-2 to remove the collagenase, oocytes are incubated with a 1:4 SOS:OR-2 solution to gradually increase Ca2+ and osmolarity (SOS in mM: 100 NaCl, 2 KCl, 1.8 CaCl2, 1 MgC12, and 5 HEPES, pH 7, with 100 U/ml penicillin, 100 µg/ml streptomycin, and 50 µg/ml gentamicin). After removing damaged and undeveloped oocytes, mature cells are incubated at 18°C for 2 h. Oocytes are then incubated with a 1:2 SOS:OR-2 solution for 30 min and finally transferred in pure SOS solution for use.

### Molecular biology

The human α1S subunits (GenBank accession no. BC133671) of skeletal CaV1.1 channels, with a Cys substituted at the extracellular flank of the S4 helix of each VSD (L159C, M519C, V893C, or S1231C for VSDs I–IV, respectively; Fig. S1 B), was coexpressed with mouse Stac3 (UniProt accession no. Q8BZ71) and rabbit β1a (UniProt accession no. P19517) in Xenopus oocytes. Single-point mutations were generated using the QuikChange Site-Directed Mutagenesis Kit (Agilent Technologies) and confirmed by sequencing. The coding RNA of the different subunits was transcribed in vitro (AmpliCap-Max T7 High Yield Message Maker Kit; CellScript) and injected into stage VI Xenopus oocytes (50 nl at 0.1–0.5 µg/µl) using a Drummond nano-injector.

### VCF

4–5 d after injection, oocytes were incubated with thiol-reactive fluorophores sensitive to environmental changes (10 µM tetramethylrhodamine-6-maleimide [TMRM-6′] for VSD-I or 20 µM MTS-5(6)-carboxytetramethylrhodamine [MTS-TAMRA] for VSDs II, III, or IV) in a depolarizing solution (in mM: 120 K-methanesulfonate [MES], 2 Ba(MES)2, and 10 HEPES, pH 7.0) for 5–20 min on ice. Subsequently, oocytes were voltage-clamped using the cut-open oocyte technique implemented for epifluorescence measurements (Stefani and Bezanilla, 1998; Pantazis and Olcese, 2013). Fluorescence changes and ionic currents were acquired simultaneously from the same membrane area (Gandhi and Olcese, 2008). The extracellular solution for ionic current recordings was (in mM) 2 Ba(MES)2, 120 NaMES, and 10 HEPES, pH 7.0. To isolate gating currents, the extracellular solution was supplemented with (in mM) 0.5 CdCl2, 0.1 LaCl3, and 0.1 ouabain to block CaV1.1 ionic currents and the Na+/K+ pump, respectively. Intracellular solution contained (in mM) 120 K-glutamate and 10 HEPES, pH 7.0. Pipette solution was (in mM) 2,700 Na-MES, 10 NaCl, and 10 Na-HEPES, pH 7.0. Prior to experiments, oocytes were injected with 100 nl of 80 mM BAPTA•4K, pH 7.0, to prevent activation of endogenous Ca2+- and Ba2+-dependent Cl channels (Barish, 1983). Holding potential was −90 mV. Ionic current and fluorescence were filtered at 1/5 of the sampling frequency (typically 1–5 kHz).

### Data analysis

The voltage dependence of ionic conductance (G(V), estimated from the peaks of the tail currents), gating charge (Q(V)), and fluorescence changes (F(V)) was characterized by fitting the data points to one Boltzmann function as F(V) = {[fmax − fmin] / [1 + exp(z · (Vm − Vhalf)(F/RT))]} + fmin, where fmax and fmin are the maximum and minimum values of the function, respectively; z is the effective valence; Vhalf is the half-activation potential; Vm is the membrane potential; T the absolute temperature; and F and R are the Faraday and Gas constants, respectively. For G(V) datasets, fmin was 0. F(V) curves were calculated at steady-state (200-ms pulses; Fig. 3 C). The F(V) curves used to compute their weighed sum were constructed from the fluorescence recording at 10 ms, which is the time of gating current integration used to construct Q(V) curves (Fig. 4 C and Fig. S1 D). Ionic current and fluorescence kinetics were fit to one or two exponential function(s). Fittings were performed by least squares using Microsoft Excel. Data are presented as mean ± SEM.

### Allosteric model of CaV activation and statistical analysis

CaV1.1 kinetics and steady-state data were described by an allosteric model of five particles: four nonidentical voltage sensors (with voltage-dependent equilibria J1...J4), coupled to a central pore that was allowed to be intrinsically voltage-dependent (equilibrium L; Fig. 6), as previously described (Pantazis et al., 2014; Savalli et al., 2016). VSDs and the pore interact with each other via coupling energies (W1...W4).

The equilibrium constant of each VSD is described by
$Ji=exp[qiekT(V−Vi)],$
where qi corresponds to the effective valance, Vi to the voltage at which Ji = 1, T the absolute temperature, k the Boltzmann constant, and e the charge of an electron. Similarly, the intrinsic equilibrium constant of the pore is described by
$L=exp[qLekT(V−VL)].$
The partition function approach was used (Sigg, 2013) to extract the steady-state voltage dependence of each particle. The partition function governing channel gating is given by Z = Zc + LZo, where
corresponds to the contracted partition function for the 16 closed states, while
corresponds to the contracted partition function for the 16 open sates. D1...D4 are related to interaction energies Wi through Di = exp(−Wi/kT).
The equilibrium curves for the five gating particles are derived from the partition function through the relations 〈k〉 = ∂ ln Z / ∂ ln K, which were used to simulate the experimental conductance (G(V)) and fluorescence (F(V)) curves. For kinetic modeling, the forward (α) and backward (β) rate constants for each transition between state n and any accessible state m are expressed as
$αn,m=νi(ZnZm)xi$
and
$βn,m=νi(ZnZm)1−xi,$
where i refers to the transitioning particle connecting states n and m, and Zn and Zm are the configuration-specific contributions to the overall partition function Z. The transition barrier between states n,m is defined by a frequency factor νi and a fractional position xi. Time evolution of all states is solved by integrating S(t) = S(0) ⋅ exp (Qt). Q is the standard rate matrix (Colquhoun and Hawkes, 1981). The initial conditions S(0) for the holding potential were obtained from S(t → ∞). The time dependence of a quantity of interest A (ionic current or fluorescence) was obtained from
$〈A(t)〉=N∑nanSn(t),$
where an is the value of the desired quantity of state n. To find the set of parameters (θ) that best reproduced all the experimental data simultaneously, we sampled the parameters space by the Markov chain Monte Carlo method using the Metropolis–Hasting algorithm as previously (Savalli et al., 2016). Briefly, posterior probabilities distributions of θ were built selecting the one that satisfies
as in Li (2012). The error function on trial i (err(θi) was as follows:
$errθi=∑ajexpj-simjθi2.$
Experiment type $(expj)$ refers to G(V) for j = 1, F(V) of VSD I–IV for j = 2..5, Im(t) for j = 6, F(V,t) of VSD I–IV for j = 7..10, and the corresponding calculated function simj generated by the model with parameters θi of ith trial in the Markov Chain. The set θi is included in the chain with probability α, as described in Li (2012).

### Online supplemental material

Fig. S1 shows topology of CaV1.1 channels, S4 segments sequence alignment, and voltage dependence of CaV1.1 channel opening for WT and Cys mutants. Fig. S2 shows that the VSD-I time course of activation follows ionic current kinetics. Fig. S3 shows that the R174W mutation disfavors CaV1.1 channel activation. Fig. S4 shows optically tracked voltage-dependent rearrangements of the individual VSD in human CaV1.1 R174W mutant. Fig. S5 shows the model prediction of each VSD contribution to pore opening. Table S1 shows fitting parameters and confidence intervals (CIs) for the allosteric model predictions. The parameters that have changed in the presence of the R174W mutation are in bold.

### Depolarization-driven structural rearrangements of each VSD in the human CaV1.1 channel exhibit distinct time- and voltage-dependent properties

To understand how CaV1.1 voltage-sensing machinery can control both the CaV1.1 pore itself and the RYR1-induced Ca2+ release, we optically tracked the structural rearrangements of each of the four VSDs of this CaV channel (Fig. 2 A), using the VCF technique, a hybrid electrophysiological–optical approach (see VCF; Mannuzzu et al., 1996; Gandhi and Olcese, 2008; Pantazis and Olcese, 2013). We engineered each VSD of the human pore-forming α1S subunit for site-directed fluorescent labeling by substituting amino acids at strategic positions in the extracellular end of the S4 helix with a cysteine suitable for labeling with thiol-reactive fluorophores (Fig. S1 B; see Molecular biology for Cys positions). Among the four helices forming each VSD, S4 segments carry a signature motif of positively charged residues able to sense changes of membrane potential and rearrange in response to this electrical stimulus (Catterall, 2010; Wu et al., 2016). The Cys constructs chosen for this study did not perturb channel function: the voltage dependence of channel activation (G(V)) and of Q(V) of Cys mutant channels appeared identical to that of WT channels (Fig. S1, C and D; Table 1; and Table 2).

We expressed human CaV1.1 channels in Xenopus oocytes (Wu et al., 2018). Oocytes were voltage-clamped using the cut-open oocyte Vaseline gap technique (Stefani and Bezanilla, 1998; Pantazis and Olcese, 2013): we simultaneously recorded membrane current and fluorescence emission from channels labeled with thiol-reactive fluorophores, used as optical reporters of VSD structural rearrangements. Voltage-dependent fluorescence signals were not detected in CaV1.1 WT channels (no engineered cysteines but 13 endogenous extracellular cysteines are present; Fig. 2 B). Also, fluorescent labeling of endogenous cysteines had no effect on the voltage dependence of channel activation (Fig. S1 C). Notably, when engineered Cys were fluorescently labeled one at the time, we resolved distinct voltage-dependent conformational changes of each CaV1.1 voltage sensor (Fig. 3 A). Fluorescence signals from VSD-I and VSD-III decreased upon depolarization, implying that fluorophores conjugated to these VSDs experienced a more quenching environment in their active state, while fluorophores conjugated to VSD-II and VSD-IV were brighter during depolarization. Overall, these results support the premise that fluorescence changes from channels with introduced Cys report VSD-specific structural rearrangements.

Each voltage sensor exhibited distinct kinetic properties that could be well-described by the sum of two exponential functions: VSD-I time-dependent properties were overall similar to ionic current activation kinetics, while VSDs II–IV displayed a faster activation component, compatible with SR Ca2+ release kinetics (Fig. 3 B and Fig. S2). Nevertheless, all VSDs exhibited a slow kinetic component of τ ≈ 100 ms (Fig. S2 A), which is prevalent in VSD-I and less represented in VSD-II, VSD-III, and VSD-IV (Fig. S2 B). Thus, using VCF, we demonstrated the existence of fast- and slow-activating CaV1.1 VSDs by directly measuring their activation kinetics, a result that confirms the predictions of Feldmeyer et al. (1990) based on ionic current recordings and those of the Flucher laboratory (Fernández-Quintero et al., 2021) based on molecular dynamics simulations, which estimated slow and fast activation kinetics for VSD-I and VSD-IV, respectively.

The voltage dependence of activation for each CaV1.1 VSD (from experiments as in Fig. 3 A) was inferred by plotting the change in fluorescence intensity (ΔF) at 200 ms against the corresponding membrane potentials ((F(V) curve, Fig. 3 C): each VSD of human CaV1.1 channels exhibited distinct voltage-dependent activation curves, all preceding the G(V) curve (voltage dependence of channel opening) on the voltage axis (Fig. 4): VSD-I was the most sensitive to changes in membrane potential (effective valence z = 3.1 ± 0.05 e0, n = 6), while VSD-IV was the least (z = 0.68 ± 0.01 e0, n = 6). Moreover, the half-activation potential (Vhalf) of the four F(V) curves spanned a range of ∼40 mV along the membrane potential axis: VSD-I was the most depolarized (Vhalf = 4.2 ± 0.8 mV, n = 6), followed by VSD-II, -III, and -IV (Fig. 3 C and Table 3).

These results reveal the striking functional diversity of the four VSDs in human CaV1.1 channels in terms of time and voltage dependence, possibly underlying distinct physiological roles.

### Charge movement and fluorescence signals arise from the same conformational changes

Before VCF was developed, gating current recordings represented the gold standard measurement to extrapolate information about the properties of voltage sensors (Bezanilla and Stefani, 1998). However, gating currents are ensemble composite signals generated by the movement of the four charged voltage sensors, and thus the computed voltage dependence of the charge movement (Q(V) curve) blends together information from all four heterologous VSDs. The Q(V) curve of human CaV1.1 channels falls within the voltage range of the four F(V) curves describing the voltage dependence of each distinct VSD (Fig. 4). As charge movement (Q(V)) and fluorescence signals (F(V)s) are expected to arise from the same voltage-dependent conformational rearrangements, a reasonable prediction is that the sum of four F(V) curves weighed by their respective effective valence (z) would approximate the Q(V) curve. For this purpose, the four F(V) curves were constructed from the amplitude of fluorescence signals after 10-ms depolarizations, which is the time of gating current integration to construct Q(V) curves. The result of this operation is reported in Fig. 4 C, which shows that Q(V) and F(V) weighed sum have practically identical voltage dependence: in other words, VCF allowed decomposition of the Q(V) curve into its four components, providing previously unattainable information.

### A naturally occurring, charge-neutralizing mutation (R174W) impairs VSD-I activation and voltage-dependent channel opening

The relevance of VSD-I to CaV1.1 activation, highlighted by the compatible kinetics (Fig. 3 B and Fig. S2), was confirmed by introducing a naturally occurring mutation that neutralizes the innermost charge of VSD-I S4 helix (R174W) and that has been linked to a skeletal muscle disorder called malignant hyperthermia (MH; Eltit et al., 2012; Bannister and Beam, 2013b). We expressed either WT or mutant CaV1.1 channels in Xenopus oocytes and measured ionic currents (Fig. S3). The activation voltage dependence of R174W channels shifted to more depolarized membrane potentials by ∼35 mV as compared with WT channels, impairing Ca2+ influx at physiological membrane potentials (Fig. 5 A). To gain a mechanistic understanding of how R174W impairs CaV1.1 pore activation, we optically tracked the movement of VSD-I using the VCF technique (Fig. S4 A). We found that the MH-linked mutation caused a ∼30 mV rightward-shift of VSD-I voltage-dependent activation and reduced the sensitivity to changes in membrane potentials (effective charge, z; Fig. 5 B and Table 3). These results demonstrate that R174 is indeed a voltage-sensing charge and further support the role of VSD-I as the main driver of pore opening.

Moreover, we found that the voltage- and time-dependent properties of VSDs II, III, and IV were not affected by the R174W mutation (Fig. 5 B and Table 3), ruling out the possibility of significant cooperativity among voltage sensors.

### VSD-I contributes the most energy to stabilizing WT CaV1.1 open states

To quantify the energetic contribution of the individual CaV1.1 VSDs to channel opening, we used a structurally relevant, allosteric model of CaV activation, previously developed for CaV1.2 channels (Pantazis et al., 2014; Savalli et al., 2016).

This model accounts for both time- and voltage-dependent properties of CaV channels and comprises five two-state gating particles, resulting in 25 = 32 possible states (16 closed states and 16 open states): a pore, which can exist either in the open or closed state, and four voltage sensors, each transitioning between resting and active states (Fig. 6; Pantazis et al., 2014; Savalli et al., 2016). The voltage-dependent equilibrium of each particle was characterized by Vi, the voltage at which they are equally distributed among the two states, and qi, the apparent number of elementary charges displaced during activation (see Allosteric model of CaV... for further details). This allosteric model has no a priori assumptions, i.e., the channel can potentially open from any of the 16 closed states. The activation of each VSD contributes some energy (Wi) to the stabilization of the open state of the pore. We performed a global simultaneous fit of ionic currents and fluorescence changes (which report the movements of each VSD) at four different membrane potentials together with their corresponding steady-state curves G(V) and F(V)s (voltage dependence of activation of the pore and VSDs, respectively) for WT and R174W channels. As shown in Fig. 7, we found a parameter set that faithfully reproduced the experimental data obtained with WT as well as R174W CaV1.1 channels simultaneously; only parameters pertaining to VSD-I were allowed to differ between WT and mutant channels (Table S1). This solution was found by Markov chain Monte Carlo sampling with 5 ∙ 106 samples in three replicas that also tested for uniqueness and provided CIs for all parameters (Fig. 7 and Table S1).

This analysis revealed that VSD-I contributes the most energy to stabilizing the open state in WT channels (−99 < W1WT < −71 meV; 95% CI; Table S1). The other three VSDs were predicted to have a smaller or negligible contribution (Wi > −21 meV), as illustrated in Fig. S5, which compares the impact of each VSD on the facilitation of voltage dependence of activation (i.e., leftward shift of G(V) curve with respect to the intrinsic voltage dependence of the pore). These plots show that the largest contribution is conferred by VSD-I (Fig. S5 A), with a smaller role of VSD-IV (Fig. S5 D) and negligible influence of VSD-II and VSD-III (Fig. S5, B and C).

Sampling the parameters space by Markov chain Monte Carlo as in Savalli et al. (2016) yields the probabilities distributions for the parameters describing the data according to the allosteric model. These distributions, depicted as violin plots in Fig. 8, indicate that the likelihood that VSD-II, VSD-III, or VSD-IV contribute to pore opening >1 kT (i.e., Wi < −25 meV) is <0.00075, <0.03, and <0.15, respectively.

Analyzing VCF data with an allosteric model for CaV activation, we also found that a single charge-neutralizing mutation in VSD-I not only made this sensor reluctant to activate but also drastically decreased the energetic coupling between VSD-I and the pore (−42 < W1R174W < −10 meV; 95% CI; Table S1). In other words, in R174W channels, when VSD-I activates, its contribution to the stabilization of the open configuration of the channel is weaker, resulting in reduced Ca2+ influx at physiological membrane potentials. Also, the valence (q1) of the perturbed VSD-I was reduced by ∼60%, consistent with the neutralization of a relevant gating charge. The fitting parameter set and relative CIs are presented in Table S1.

The allosteric model developed here predicts that the pore has an intrinsic voltage dependence of ∼1.7 e0 (1.6 < qL < 1.8 e0; 95% CI). A salient feature of allosteric interactions is that channel opening will still occur when all the sensors are at the resting state. However, if the pore domain conformational changes during channel gating are intrinsically voltage-dependent, a continuous increase in the channel Popen should be observed after all voltage sensors are maximally active. This behavior can be visualized by plotting the G(V) as a Hill plot adapted for voltage-dependent equilibrium ln[G(V)/(1 − G(V)] (Sigg, 2013). With such transformation, a two-state system yields a straight line with a slope corresponding to the effective valence and intercept (G(V) = 0.5) at V1/2. The Hill-transformed G(V) for both WT and R174W channels became asymptotic for positive membrane potentials with a slope representing the effective valence of the intrinsic voltage dependence of the pore (i.e., without the contribution of the VSDs; Sigg, 2013). Note that the model approximates the data well, yielding the intrinsic effective valence of the pore (qL = 1.7 e0). The allosteric model also predicted that R174W mutation is associated with a reduction of VSD–pore coupling energy of 59 meV (ΔΔG = ∑WWT− ∑WR174W) (Fig. 9).

In summary, the 32-state allosteric model of CaV gating lends further support to a predominant role of VSD-I in CaV1.1 activation and assigned a smaller contribution of VSD-IV. Based on this analysis, VSD-II and VSD-III would not contribute to channel opening. We speculate that VSD-II and VSD-III remain available for other voltage-dependent channel functions, such as RYR1 activation and, consequently, SR Ca2+ release, since EC coupling remains functional in patients carrying a R174W missense mutation of CaV1.1 (Eltit et al., 2012).

Using VCF and computational modeling, we have dissected the operation of the voltage-sensing apparatus of the human skeletal muscle CaV1.1 channel, revealing the functional properties of each of its four voltage sensors. The experimental results expose a striking heterogeneity in their voltage- and time-dependent properties and shed light on the apparent paradox whereby an extremely slowly activating channel controls a very fast process, SR Ca2+ release, that takes place during 2–5 ms of the skeletal muscle action potential. We found that VSD-I plays a major role in CaV1.1 channel activation, based on several pieces of evidence: (1) VSD-I time-dependent properties are compatible with ionic current activation kinetics (Fig. 3 B and Fig. S2); (2) the neutralization of the innermost charge of VSD-I (R174W) impairs the activation of this voltage sensor, in turn compromising voltage-dependent CaV1.1 pore opening at physiological membrane potentials (Fig. 5); and (3) a statistical mechanical model relevant to CaV molecular architecture predicts that the open state of the human CaV1.1 pore is mostly stabilized by VSD-I activation (W1 = −81 meV or ∼3 kT) with a minimal contribution of the other VSDs (W2-4 > −25 meV or <1 kT; Fig. 7 and Table S1). Other mutations of voltage-sensing charges in VSD-I also affect ionic current voltage dependence (García et al., 1997; El Ghaleb et al., 2019), in agreement with a role of VSD-I in channel opening. On the other hand, VSD-II, -III, and -IV display activation components ∼20-fold faster than VSD-I (Fig. 3 B and Fig. S2), a feature that qualifies them as kinetically compatible candidates for a role as voltage sensor(s) of SR Ca2+ release. Since a large body of evidence implicates the intracellular II–III loop for a major role in transducing depolarization sensed by CaV1.1 to RYR1 opening (Bannister and Beam, 2013a), we speculate that either VSD-II or VSD-III (or both) represent the voltage sensor(s) of skeletal EC coupling (Fig. 10).

### CaV1.1 channels exhibit a unique mechanism for voltage-dependent channel opening, among the super family of voltage-gated ion channels

The present study suggests that CaV1.1 complexes operate mainly upon activation of a single voltage sensor (VSD-I), a feature which has not been observed in any other voltage-gated ion channel (Smith-Maxwell et al., 1998; Chanda and Bezanilla, 2002; Pantazis et al., 2014). Indeed, KV channels typically require activation of all four VSDs to favor pore opening (Smith-Maxwell et al., 1998), while NaV channels require the activation of three VSDs (specifically, VSDs I, II, and III; Chanda and Bezanilla, 2002; Goldschen-Ohm et al., 2013). In addition, we recently discovered that pore opening in CaV1.2 channels is mainly driven by two voltage sensors (VSDs II and III; Pantazis et al., 2014; Savalli et al., 2016).

Previous studies in rabbit embryonic and adult CaV1.1 variants suggested that VSD-IV controls voltage sensitivity of pore opening (Tuluc et al., 2016). However, in the human adult CaV1.1 channel used in our study, the computational analysis with a CaV allosteric model predicts a small energetic contribution of VSD-IV activation to the stabilization of the pore open state (Table S1). The difference with previous studies regarding the role of VSD-IV may arise from the different CaV1.1 isoform (human in this study versus rabbit in Tuluc et al. [2016]) or from the different cellular environment (Xenopus oocytes in this study versus mouse dysgenic myotubes in Tuluc et al. [2016]). It is also possible that the deletion of 19 amino acids in VSD-IV of embryonic channels increases the energetic contribution of this sensor to channel activation.

### Different L-type Ca2+ channels engage different VSDs to open their pores

In skeletal CaV1.1 channels composed of α1S1a, the mid-point of VSDs activation spans a range of ∼40 mV (Table 3), while in CaV1.2 channels (α1C3), the extent is larger (∼80 mV; Savalli et al., 2016), implying a more loose coupling between voltage sensors and pore in the latter case. Indeed, the overall energetic coupling (∑W) between VSDs and pore quantified by an allosteric model of CaV channel activation supports this view: in skeletal channels, ∑W is ∼100 meV (Table S1), while in CaV1.2 channels, ∑W corresponds to a smaller value (∼40 meV; Savalli et al., 2016). Thus, the mechanism of voltage-dependent pore opening in L-type CaV channels appears to be isoform-specific and does not conform to a single conserved mechanism, despite ∼90% sequence homology in the transmembrane segments and practically identical S4 helices. Conceivably, the difference in extracellular and intracellular linkers mediates the large functional difference of CaV1.1 and CaV1.2 channels, resulting in very different physiological functions. Indeed, previous studies in CaV channels (Nakai et al., 1994; Tuluc et al., 2016; Liu et al., 2017) and other voltage-gated ion channels (Priest et al., 2013) demonstrate the importance of the S3–S4 linker sequence and length in defining the biophysical properties of channel activation.

### Conclusions and perspectives

We have pioneered VCF in human CaV1.1 complexes and resolved unique time- and voltage-dependent properties of its four VSDs. We explained how CaV1.1 channels solve the task of controlling both the fast (1–3 ms) SR Ca2+ release via RYR1 activation and the slow (40–60 ms) opening of its own pore, by assigning each function to distinct VSDs. VSD-I qualifies as the main contributor of CaV1.1 activation, a new solution for voltage-dependent channel activation, while VSD-II and -III exhibit biophysical properties compatible with the timing of SR Ca2+ release, likely representing the voltage sensor(s) of skeletal EC-coupling mechanism. More studies are certainly needed to reveal the precise role of Cav1.1 VSDs in RYR1 gating and EC coupling. However, the experimental strategy developed here sets a framework to mechanistically address clinically relevant questions on the functional consequences of disease-causing mutations in CaV1.1 voltage dependence (e.g., hypokalemic periodic paralysis and MH-associated mutations; Flucher, 2020). The approach developed in this work can also advance our knowledge on the fundamental mechanisms underlying CaV1.1 modulation by auxiliary subunits or alternative splicing, benefitting from the new level of information attainable by simultaneously acquiring structural data from functional channel proteins operating in a membrane environment.

Eduardo Ríos served as editor.

We are thankful for the insightful discussions and comments of Dr. Michela Ottolia and Dr. Andreas Schwingshackl and their laboratory members. We thank Dr. Daniel Sigg for insightful discussions on the mathematical modeling.

This work was supported by the National Institutes of Health/National Institute of General Medical Sciences (grant R35GM131896 to R. Olcese), the National Institutes of Health/National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant R01AR063182 to S.C. Cannon); an American Heart Association Postdoctoral Fellowship (17POST33670046 to M. Angelini); and Chilean Government (grants FONDECYT 1161672 to A. Neely). The Centro Interdisciplinario de Neurociencia de Valparaíso is a Millennium Institute supported by the Millennium Scientific Initiative of the Chilean Ministry of Economy.

The authors declare no competing financial interests.

Author contributions: Conceptualization, N. Savalli, A. Neely, S.C. Cannon, and R. Olcese; Investigation, N. Savalli, M. Angelini, F. Steccanella, J. Wier, F. Wu, M. Quinonez, and M. DiFranco; Software, N. Savalli, A. Neely, and R. Olcese; Formal Analysis, N. Savalli, M. Angelini, F. Steccanella, A. Neely, and R. Olcese; Writing—Original Draft, N. Savalli and R. Olcese; Writing—Review & Editing, N. Savalli, M. Angelini, F. Steccanella, M. DiFranco, A. Neely, S.C. Cannon, and R. Olcese; Funding Acquisition, M. Angelini, A. Neely, S.C. Cannon, and R. Olcese.

Bannister
,
R.A.
2016
.
Bridging the myoplasmic gap II: more recent advances in skeletal muscle excitation-contraction coupling
.
J. Exp. Biol.
219
:
175
182
.
Bannister
,
R.A.
, and
K.G.
Beam
.
2013
a
.
Ca(V)1.1: The atypical prototypical voltage-gated Ca2+ channel
.
Biochim. Biophys. Acta.
1828
:
1587
1597
.
Bannister
,
R.A.
, and
K.G.
Beam
.
2013
b
.
Impaired gating of an L-Type Ca(2+) channel carrying a mutation linked to malignant hyperthermia
.
Biophys. J.
104
:
1917
1922
.
Barish
,
M.E.
1983
.
A transient calcium-dependent chloride current in the immature Xenopus oocyte
.
J. Physiol.
342
:
309
325
.
Bezanilla
,
F.
, and
E.
Stefani
.
1998
.
Gating currents
.
Methods Enzymol.
293
:
331
352
.
Catterall
,
W.A.
2010
.
Ion channel voltage sensors: structure, function, and pathophysiology
.
Neuron.
67
:
915
928
.
Catterall
,
W.A.
,
M.J.
Lenaeus
, and
T.M.
Gamal El-Din
.
2020
.
Structure and Pharmacology of Voltage-Gated Sodium and Calcium Channels
.
Annu. Rev. Pharmacol. Toxicol.
60
:
133
154
.
Cha
,
A.
,
P.C.
Ruben
,
A.L.
George
Jr
.,
E.
Fujimoto
, and
F.
Bezanilla
.
1999
.
Voltage sensors in domains III and IV, but not I and II, are immobilized by Na+ channel fast inactivation
.
Neuron.
22
:
73
87
.
Chanda
,
B.
, and
F.
Bezanilla
.
2002
.
Tracking voltage-dependent conformational changes in skeletal muscle sodium channel during activation
.
J. Gen. Physiol.
120
:
629
645
.
Colquhoun
,
D.
, and
A.G.
Hawkes
.
1981
.
On the stochastic properties of single ion channels
.
Proc. R. Soc. Lond. B Biol. Sci.
211
:
205
235
.
Dolphin
,
A.C.
2016
.
Voltage-gated calcium channels and their auxiliary subunits: physiology and pathophysiology and pharmacology
.
J. Physiol.
594
:
5369
5390
.
El Ghaleb
,
Y.
,
M.
Campiglio
, and
B.E.
Flucher
.
2019
.
Correcting the R165K substitution in the first voltage-sensor of CaV1.1 right-shifts the voltage-dependence of skeletal muscle calcium channel activation
.
Channels (Austin).
13
:
62
71
.
Eltit
,
J.M.
,
R.A.
Bannister
,
O.
Moua
,
F.
Altamirano
,
P.M.
Hopkins
,
I.N.
Pessah
,
T.F.
Molinski
,
J.R.
López
,
K.G.
Beam
, and
P.D.
Allen
.
2012
.
Malignant hyperthermia susceptibility arising from altered resting coupling between the skeletal muscle L-type Ca2+ channel and the type 1 ryanodine receptor
.
109
:
7923
7928
.
Feldmeyer
,
D.
,
W.
Melzer
,
B.
Pohl
, and
P.
Zöllner
.
1990
.
Fast gating kinetics of the slow Ca2+ current in cut skeletal muscle fibres of the frog
.
J. Physiol.
425
:
347
367
.
Fernández-Quintero
,
M.L.
,
Y.
El Ghaleb
,
P.
Tuluc
,
M.
Campiglio
,
K.R.
Liedl
, and
B.E.
Flucher
.
2021
.
Structural determinants of voltage-gating properties in calcium channels
.
eLife.
10
:e64087.
Flucher
,
B.E.
2020
.
Skeletal muscle CaV1.1 channelopathies
.
Pflugers Arch.
472
:
739
754
.
Flucher
,
B.E.
, and
M.
Campiglio
.
2019
.
STAC proteins: The missing link in skeletal muscle EC coupling and new regulators of calcium channel function
.
Biochim. Biophys. Acta Mol. Cell Res.
1866
:
1101
1110
.
Gandhi
,
C.S.
, and
R.
Olcese
.
2008
.
The voltage-clamp fluorometry technique
.
Methods Mol. Biol.
491
:
213
231
.
García
,
J.
,
J.
Nakai
,
K.
Imoto
, and
K.G.
Beam
.
1997
.
Role of S4 segments and the leucine heptad motif in the activation of an L-type calcium channel
.
Biophys. J.
72
:
2515
2523
.
Goldschen-Ohm
,
M.P.
,
D.L.
Capes
,
K.M.
Oelstrom
, and
B.
Chanda
.
2013
.
Multiple pore conformations driven by asynchronous movements of voltage sensors in a eukaryotic sodium channel
.
Nat. Commun.
4
:
1350
.
Hering
,
S.
,
E.M.
Zangerl-Plessl
,
S.
Beyl
,
A.
Hohaus
,
S.
Andranovits
, and
E.N.
Timin
.
2018
.
Calcium channel gating
.
Pflugers Arch.
470
:
1291
1309
.
Hernández-Ochoa
,
E.O.
, and
M.F.
Schneider
.
2018
.
Voltage sensing mechanism in skeletal muscle excitation-contraction coupling: coming of age or midlife crisis?
Skelet. Muscle.
8
:
22
.
Hoffmann
,
H.
2021
.
Violin Plot
. MATLAB Central File Exchange. https://www.mathworks.com/matlabcentral/fileexchange/45134-violin-plot (Accessed September 14, 2021).
Li
,
Y.
2012
.
MOMCMC: An efficient Monte Carlo method for multi-objective sampling over real parameter space
.
Comput. Math. Appl.
64
:
3542
3556
.
Liu
,
N.
,
Y.
Liu
,
Y.
Yang
, and
X.
Liu
.
2017
.
Linker flexibility of IVS3-S4 loops modulates voltage-dependent activation of L-type Ca2+ channels
.
Channels (Austin).
11
:
34
45
.
Mannuzzu
,
L.M.
,
M.M.
Moronne
, and
E.Y.
Isacoff
.
1996
.
Direct physical measure of conformational rearrangement underlying potassium channel gating
.
Science.
271
:
213
216
.
Nakai
,
J.
,
B.A.
,
K.
Imoto
, and
K.G.
Beam
.
1994
.
Critical roles of the S3 segment and S3-S4 linker of repeat I in activation of L-type calcium channels
.
91
:
1014
1018
.
Niu
,
J.
,
W.
Yang
,
D.T.
Yue
,
T.
Inoue
, and
M.
Ben-Johny
.
2018
.
Duplex signaling by CaM and Stac3 enhances CaV1.1 function and provides insights into congenital myopathy
.
J. Gen. Physiol.
150
:
1145
1161
.
Pantazis
,
A.
, and
R.
Olcese
.
2013
.
Cut-open Oocyte Voltage Clamp Technique
. In
Encyclopedia of Biophysics.
G.C.K.
Roberts
, editor.
Springer
,
Berlin, Heidelberg
.
406
413
.
Pantazis
,
A.
,
N.
Savalli
,
D.
Sigg
,
A.
Neely
, and
R.
Olcese
.
2014
.
Functional heterogeneity of the four voltage sensors of a human L-type calcium channel
.
111
:
18381
18386
.
Polster
,
A.
,
S.
Perni
,
H.
Bichraoui
, and
K.G.
Beam
.
2015
.
Stac adaptor proteins regulate trafficking and function of muscle and neuronal L-type Ca2+ channels
.
112
:
602
606
.
Priest
,
M.F.
,
J.J.
Lacroix
,
C.A.
Villalba-Galea
, and
F.
Bezanilla
.
2013
.
S3-S4 linker length modulates the relaxed state of a voltage-gated potassium channel
.
Biophys. J.
105
:
2312
2322
.
Rios
,
E.
, and
G.
Brum
.
1987
.
Involvement of dihydropyridine receptors in excitation-contraction coupling in skeletal muscle
.
Nature.
325
:
717
720
.
Ríos
,
E.
,
G.
Pizarro
, and
E.
Stefani
.
1992
.
Charge movement and the nature of signal transduction in skeletal muscle excitation-contraction coupling
.
Annu. Rev. Physiol.
54
:
109
133
.
Savalli
,
N.
,
A.
Pantazis
,
D.
Sigg
,
J.N.
Weiss
,
A.
Neely
, and
R.
Olcese
.
2016
.
The α2δ-1 subunit remodels CaV1.2 voltage sensors and allows Ca2+ influx at physiological membrane potentials
.
J. Gen. Physiol.
148
:
147
159
.
Schneider
,
M.F.
, and
W.K.
Chandler
.
1973
.
Voltage dependent charge movement of skeletal muscle: a possible step in excitation-contraction coupling
.
Nature.
242
:
244
246
.
Shishmarev
,
D.
2020
.
Excitation-contraction coupling in skeletal muscle: recent progress and unanswered questions
.
Biophys. Rev.
12
:
143
153
.
Sigg
,
D.
2013
.
A linkage analysis toolkit for studying allosteric networks in ion channels
.
J. Gen. Physiol.
141
:
29
60
.
Smith-Maxwell
,
C.J.
,
J.L.
Ledwell
, and
R.W.
Aldrich
.
1998
.
Role of the S4 in cooperativity of voltage-dependent potassium channel activation
.
J. Gen. Physiol.
111
:
399
420
.
Stefani
,
E.
, and
F.
Bezanilla
.
1998
.
Cut-open oocyte voltage-clamp technique
.
Methods Enzymol.
293
:
300
318
.
Tuluc
,
P.
,
B.
Benedetti
,
P.
Coste de Bagneaux
,
M.
Grabner
, and
B.E.
Flucher
.
2016
.
Two distinct voltage-sensing domains control voltage sensitivity and kinetics of current activation in CaV1.1 calcium channels
.
J. Gen. Physiol.
147
:
437
449
.
Varga
,
Z.
,
W.
Zhu
,
A.R.
Schubert
,
J.L.
Pardieck
,
A.
Krumholz
,
E.J.
Hsu
,
M.A.
Zaydman
,
J.
Cui
, and
J.R.
Silva
.
2015
.
Direct Measurement of Cardiac Na+ Channel Conformations Reveals Molecular Pathologies of Inherited Mutations
.
Circ. Arrhythm. Electrophysiol.
8
:
1228
1239
.
Wu
,
J.
,
Z.
Yan
,
Z.
Li
,
X.
Qian
,
S.
Lu
,
M.
Dong
,
Q.
Zhou
, and
N.
Yan
.
2016
.
Structure of the voltage-gated calcium channel Ca(v)1.1 at 3.6 Å resolution
.
Nature.
537
:
191
196
.
Wu
,
F.
,
M.
Quinonez
,
M.
DiFranco
, and
S.C.
Cannon
.
2018
.
Stac3 enhances expression of human CaV1.1 in Xenopus oocytes and reveals gating pore currents in HypoPP mutant channels
.
J. Gen. Physiol.
150
:
475
489
.
Zhao
,
Y.
,
G.
Huang
,
J.
Wu
,
Q.
Wu
,
S.
Gao
,
Z.
Yan
,
J.
Lei
, and
N.
Yan
.
2019
.
Molecular Basis for Ligand Modulation of a Mammalian Voltage-Gated Ca2+ Channel
.
Cell.
177
:
1495
1506.e12
.

This work is part of a special issue on excitation–contraction coupling.