Lobe-related concentration- and Ca²⁺-dependent interactions of calmodulin with C- and N-terminal tails of the Ca_V1.2 channel

Abstract

This study examined the bindings of calmodulin (CaM) and its mutants with the C- and N-terminal tails of the voltage-gated Ca²⁺ channel Ca_V1.2 at different CaM and Ca²⁺ concentrations ([Ca²⁺]) by using the pull-down assay method to obtain basic information on the binding mode, including its concentration- and Ca²⁺-dependencies. Our data show that more than one CaM molecule could bind to the Ca_V1.2 C-terminal tail at high [Ca²⁺]. Additionally, the C-lobe of CaM is highly critical in sensing the change of [Ca²⁺] in its binding to the C-terminal tail of Ca_V1.2, and the binding between CaM and the N-terminal tail of Ca_V1.2 requires high [Ca²⁺]. Our data provide new details on the interactions between CaM and the Ca_V1.2 channel.

Introduction

Voltage-gated Ca²⁺ channels (Ca_Vs) control cellular Ca²⁺ entry in response to changes in the membrane potential, and play essential roles in the generation of cardiac action potentials, excitation–contraction coupling, hormone and neurotransmitter release, and activity-dependent gene transcription [1]. Ca²⁺ is not only a potent activator of intracellular signaling pathways but also a toxicant when overloaded [2]. As the major routes of Ca²⁺ influx, the Ca_Vs are regulated by two self-regulatory Ca²⁺-dependent feedback mechanisms, Ca²⁺-dependent facilitation (CDF) and Ca²⁺-dependent inactivation (CDI) [1, 3]. CDF is an increase in the channel opening when basal Ca²⁺ level is moderately increased or repeated transient depolarization occurs, which may augment the Ca²⁺ signal, whereas CDI is an enhancement of the channel closing when the entry of Ca²⁺ is further increased, which may prevent the harmful actions of Ca²⁺ overload [1]. Calmodulin (CaM) is a ubiquitously expressed Ca²⁺-binding protein [2], and it is believed that both CDF and CDI require CaM to bind with the α₁ subunit of Ca_Vs [3]. Especially, it has recently been demonstrated that the binding between CaM and the Ca_V1.2 channel and the induced Ca²⁺-dependent regulations play very important roles not only in physiological functions but also in pathophysiological conditions [4].

Numerous studies have suggested that CaM is constitutively tethered to the channel, and that the interaction of CaM with the proximal part of the C-terminal tail of the channel plays a prominent role in CDF and CDI [5–8]. The tethering or binding sites in the C-terminal tail of Ca_V1.2 include an isoleucine–glutamine motif [5–7] (referred as the IQ motif, a consensus CaM-binding motif in many proteins) and a preIQ motif (also referred as peptide A and C) [5, 8] (Fig. 1a). The location of the preIQ motif is between the IQ motif and the Ca²⁺-binding helix-loop-helix structural motif called ‘EF-hand motif’ in the C-terminal tail of Ca_V1.2 (Fig. 1a). Besides the CaM binding sites in the C-terminal tail of Ca_V1.2, other regions have also been defined, including the I–II loop [8] and a CaM-binding motif in the N-terminal tail of Ca_V1.2 and Ca_V1.3, named N-terminal Spatial Ca²⁺ Transforming Element (NSCaTE) [9] (Fig. 1a). However, a report has shown that CaM cannot foster interactions between the N- and C-terminal tails to promote CDI [10]. CaM may show three different Ca²⁺-binding states, namely, the Ca²⁺-free form (apoCaM), and the partially and fully Ca²⁺-bound forms (Ca²⁺/CaM: ${\text{Ca}}_{2}^{2 + } {\text{CaM}}$ and ${\text{Ca}}_{4}^{2 + } {\text{CaM}}$, respectively), which are likely to be physiologically relevant [2, 5, 6]. It is reported that apoCaM may bind to the C-terminal tail of Ca_V1.2 [5], showing that CaM may already tether to the channel even at resting condition, but it is not clear whether apoCaM may bind to the N-terminal tail of Ca_V1.2 or not. Thus, further work is needed to delineate the interaction of CaM with the N-terminal tail of Ca_V1.2 and its Ca²⁺-dependent characteristics.

Fig. 1

Schematic illustrations of the peptides of Ca_V1.2 channel, CaM and its mutants. a NT is a N-terminal peptide containing a CaM-binding site called NSCaTE (see text). The C-terminal tail was divided into three parts as illustrated. CT1 contains regulatory regions (shown as boxes), Ca²⁺-binding EF-hand motif (EF) and two CaM-binding motifs (preIQ and IQ). Amino acid (a.a.) numbers (guinea-pig cardiac Ca_V1.2) of peptides are shown in the parentheses. LI–II, LII–III and LIII–IV represent intracellular loops between homologous repeats of I to IV. b Schematic illustrations of CaM and its mutants (CaM₁₂, CaM₃₄ and CaM₁₂₃₄). Closed circles represent normal Ca²⁺-binding sites in N-lobe (N) and C-lobe (C) of CaM, open circles with X represent mutated Ca²⁺-binding sites, which are unable to bind with Ca²⁺. Substitutions of a.a. are shown with one-letter codes of a.a. c A.a. sequences of human CaM. The sequences of Ca²⁺-binding sites 1–4 and mutated points are shown in boxes and red color, respectively (color figure online)

The concentration of the intracellular free Ca²⁺ at resting condition (~10⁻⁷ M) is 10⁴ times lower than those outside the cells or in the intracellular stores (~10⁻³ M) [2]. Various extracellular stimuli promote the influx or release of Ca²⁺ from either outside the cells or the intracellular stores into the intracellular milieu. This increased intracellular Ca²⁺ is only briefly available to act as a cellular signal because Ca²⁺-binding proteins and Ca²⁺ pumps immediately combine to sequester and transport it to intracellular storage sites or outside the cell. Structurally, a CaM has two lobes, each of which contains two EF-hand conformational Ca²⁺-binding sites (Fig. 1b). The two Ca²⁺-binding sites of the C-terminal lobe (C-lobe) have 3- to 5-fold higher affinity for Ca²⁺ than those of the N-terminal lobe (N-lobe) [2]. CaM mutants, such as CaM₁₂ (mutated at Ca²⁺-binding sites 1 and 2 in the N-lobe), CaM₃₄ (mutated at sites 3 and 4 in the C-lobe), and CaM₁₂₃₄ (mutated all sites above) (Fig. 1b, c), are thought to represent different Ca²⁺-binding states of CaM, and have been used to investigate Ca²⁺-dependent regulation of Ca_V1.2 channel [5, 6]. A number of experiments have demonstrated that CDF and CDI are controlled independently by the N- and C-lobes [6, 11, 12]. However, the mechanism underlying CaM lobe-specific regulations of the Ca_V1.2 channel remains to be settled.

Another critical issue in understanding the machinery of CDF and CDI is to deduce how many CaM molecules are involved in the Ca²⁺-dependent regulations. Mori et al. [13] reported that one CaM appears to be both necessary and sufficient to produce CDI of its associated channel. However, a recent crystallographic study provided evidence for multiple CaM molecules binding to the preIQ–IQ region of Ca_V1.2 [14]. Thus, although it has been established that CaM binds to the Ca_V1.2 channel at multiple sites, it is not fully understood how CaM interacts with the channel during Ca²⁺-dependent regulation.

In the present study, we have examined the bindings of CaM and its Ca²⁺-binding site mutants to the C- and N-terminal tails of the Ca_V1.2 channel at different CaM concentrations and different [Ca²⁺] by using a semi-quantitative pull-down method. We have found that CaM bound to C-terminal tail in a Ca²⁺-dependent manner in which the C-lobe of CaM is much critical for sensing the change of Ca²⁺ concentration, and more than one molecule of CaM could bind to the Ca_V1.2 C-terminal region at high [Ca²⁺]. Our data provide new and basic information on the interaction between CaM and Ca_V1.2 channel.

Materials and methods

Preparation of GST-fusion peptides of Ca_V1.2 channel

The cDNAs corresponding to three C-terminal fragments of guinea pig Ca_V1.2 channel (CT1, a.a.1,509–1,789; CT2, a.a.1,778–2,003; CT3, a.a.1,942–2,169; GenBank AB016287) and nearly full-length N-terminal tail (NT, a.a.6–140) were cloned into the pGEX6P-3 plasmid (GE Healthcare Biosciences) (Fig. 1a). The corresponding peptides were expressed as GST-fusion proteins in Escherichia coli BL21 (DE3). Due to the limited solubility of the α_1C fragments, in particular CT1, the bacterial precipitates were rotated with 1.5 % N-lauroylsarcosine sodium salt (Sigma-Aldrich) for 30 min at 4 °C [15, 16]. The fusion proteins were purified with Glutathione Sepharose 4B beads (GS-4B; GE Healthcare Biosciences). The purified peptides were analyzed by SDS-PAGE, and quantified by Bradford assay using bovine serum albumin (BSA) as standard.

Preparation of CaM and its mutants

The plasmids of CaM were constructed with the cDNAs of human CaM cloned from HEK293 and pGEX6P-3 vector (GE Healthcare). Point mutations in the Ca²⁺-binding sites of CaM were introduced by Quickchange™ site-directed mutagenesis kit (QIAGEN). CaM and its mutants CaM₁₂, CaM₃₄, and CaM₁₂₃₄, mutated as E31A + E67A, S101F + E140A and E31A + E67A + S101F + E140A, respectively (Fig. 1b, c), were expressed and purified similarly to the peptides mentioned above. The GST regions were cleaved by PreScission protease (GE Healthcare). The purified peptides were checked by SDS-PAGE and quantified by Bradford assay using BSA as standard.

Binding assay and data analyses

The interactions of CaM and its mutants with the CT1, CT2, CT3, and NT regions of Ca_V1.2 channel were examined at different CaM concentrations (0.1, 0.35, 0.7, 1.4, 2.1, 3.5, 10, and 50 μM) and different free Ca²⁺ concentrations ([Ca²⁺] ≈ free, 100 nM, 10 μM, and 2 mM), using a semi-quantitative GST pull-down assay. The [Ca²⁺] was calculated by using WEBMAXC v2.10 software. Accordingly, the test solutions containing free, 100 nM, and 10 μM Ca²⁺ were prepared with 5 mM EGTA and 0, 4.79, 5.01 mM CaCl₂, respectively; the test solution of 2 mM Ca²⁺ was prepared without EGTA and with 2 mM CaCl₂ in Tris buffer (50 mM Tris, 150 mM NaCl, pH 8.0). GST-fusion fragments of the Ca_V1.2 channel (2–4 μg) were immobilized onto GS-4B beads and incubated with CaM or its mutants in 300 μl of Tris buffer for 4 h at 4 °C under agitation. Then, the beads were gently washed twice with the same buffer containing the non-ionic detergent Tween 20 (0.05 %, Sigma-Aldrich), as previously described [15]. Bound CaM and α_1C peptides were resuspended in SDS sample loading buffer and applied to 15 % SDS-PAGE. Proteins were visualized by Coomassie brilliant blue R (CBB) staining. Peptide amounts in SDS-PAGE gel were quantified by scanning the bands with the Photoshop software (Adobe, San Jose, CA, USA), and the optical density was analyzed by software Image J (NIH, Bethesda, MD, USA). The standard curves were generated with authentic BSA. The concentrations of CaM and GST-fusion peptides were corrected based on the relative optical densities of 0.59 and 0.80 estimated with the same amount of CaM and GST on the gel in reference to BSA, respectively [16].

Curve fitting of the total bound [CaM] was performed with the software SigmaPlot 10.0, assuming that free [CaM] in our experimental conditions is nearly equal to total [CaM]. According to the law of mass action, bound [CaM] (Y) in the one-site model is expressed by the Hill equation as follows.

$$ Y = B_{\hbox{max} } \cdot \frac{X}{{K_{\text{d}} + X}} $$

where B _max is the maximum binding, X is the concentration of the free ligand, and K _d is the apparent dissociation constant.

In the case of the two-site model, a sum of two Hill’s equations was used to assume independent binding.

$$ Y = B_{{\max_{1} }} \cdot \frac{X}{{K_{{d_{1} }} + X}} + B_{{\max_{2} }} \cdot \frac{X}{{K_{{d_{2} }} + X}} $$

where $B_{{\max_{1} }}$, $K_{{d_{1} }}$, $B_{{\max_{2} }}$, and $K_{{d_{2} }}$ denote B _max and K _d for the high and low affinity sites, respectively.

Data are presented as means ± standard errors (SE). Student’s t test was used to evaluate the statistical significance and P < 0.05 was considered significant.

Results

Concentration- and Ca²⁺-dependent binding of wild type CaM to the C-terminal tail of Ca_V1.2

We prepared three GST-fusion peptides derived from the C-terminal tail of Ca_V1.2: CT1 (proximal), CT2 (middle), and CT3 (distal) (Fig. 1a). CT1 contained the EF-hand, preIQ, and IQ motifs, with the latter two being known to interact with CaM (Fig. 1a). We investigated the bindings of CaM to the three fragments of the Ca_V1.2 at different CaM concentrations and different [Ca²⁺] by using a semi-quantitative pull-down assay method. In the control, CaM binding to GST was negligible (data not shown). CaM binding to CT1 was evident, while the binding to CT2 or CT3 was hardly detected (data not shown). These results are consistent with previous reports [5, 8, 15, 16]. Therefore, the following experiments were focused on the CT1 peptide.

The examples of CaM binding with CT1 are shown in Fig. 2a. CaM at various concentrations (0.1–50 μM) was incubated with GST-fusion CT1 peptide in the presence of Ca²⁺ ranging from free (unavailable for binding) to 2 mM. The binding of CaM to the GST-fusion CT1 peptide increased with increasing concentrations of CaM at any [Ca²⁺] including the Ca²⁺-free condition, and it increased with increasing [Ca²⁺] at fixed CaM concentration.

Fig. 2

CaM and CaM₁₂ binding to CT1 peptide of the Ca_V1.2 channel. GST pull-down assay for CaM (a) or CaM₁₂ (c) binding to CT1. Protein bands are shown as Coomassie brilliant blue (CBB) staining at different [Ca²⁺]. CaM bands are indicated by arrows. Plots of concentration-dependent binding with fitted curves for CaM (b) and CaM₁₂ (d) (0.1–50 μM) to CT1 at ≈free, 100 nM, 10 μM, and 2 mM Ca²⁺. Band densities were converted into molar ratio (CaM/GST-CT1) and plotted against total [CaM] or [CaM₁₂] with symbols as indicated in each graph and with mean ± SE (n = 3–10). *P < 0.05, **P < 0.01, compared with corresponding bindings at Ca²⁺-free condition

Figure 2b shows the plots of bound CaM against total [CaM] along with the fitted curves. The parameters for the fitted curves are shown in Table 1. In Ca²⁺-free condition and 100 nM Ca²⁺, the binding between CaM and CT1 appeared to follow the one-site model. At 10 μM and 2 mM Ca²⁺, the bindings of CaM to CT1 were better fitted by the two-site model. The maximal binding, estimated as $B_{{\max_{1} }}$ + $B_{{\max_{2} }}$, is less than 1 mol/mol CaM/peptide at free and 100 nM Ca²⁺, suggesting that one Ca_V1.2 channel peptide might bind one CaM molecule. When [Ca²⁺] was increased, the maximal binding reaches over 1 mol/mol CaM/peptide (1.02 for 10 μM, 1.74 for 2 mM Ca²⁺), suggesting that CT1 could bind more than one CaM molecule at higher [Ca²⁺] (≥10 μM). The above results demonstrated that the binding of CaM to CT1 was dependent on the concentrations of both CaM and Ca²⁺.

Table 1 Parameters for the bindings of CaM and its mutants to CT1

Concentration- and Ca²⁺-dependent binding of CaM₁₂ to CT1

The bindings of CaM₁₂ to CT1 were then examined. As shown in Fig. 2c, CaM₁₂ was able to bind CT1 in a concentration-dependent manner, and although less obvious than for the wild-type CaM, the Ca²⁺-dependency was maintained. As shown in Fig. 2d, the fitted curves shift upward, indicating that the Ca²⁺-dependency of CaM₁₂ binding to CT1 still exists. Similarly to the binding features of wild-type CaM, at free and 100 nM Ca²⁺, the one-site model is proper to describe the binding dynamics, while at 10 μM and 2 mM Ca²⁺, the two-site model is more appropriate.

Concentration- and Ca²⁺-dependent binding of CaM₃₄ and CaM₁₂₃₄ to CT1

The bindings of CaM₃₄ and CaM₁₂₃₄ were also examined. As shown in Fig. 3, these two mutated CaM were still able to bind to CT1 in a concentration-dependent manner but not in a Ca²⁺-dependent manner. The bindings of the two mutated CaM to CT1 were up to 1 mol/mol (CaMs/CT1), implying that their binding capability was not much affected (Fig. 3b, d).

Fig. 3

CaM₃₄ and CaM₁₂₃₄ binding to CT1 peptide of the Ca_V1.2 channel. GST pull-down assay for CaM₃₄ (a) or CaM₁₂₃₄ (c) binding to CT1. Protein bands are shown as CBB staining at different [Ca²⁺]. CaM bands are indicated by arrows. Plots of concentration-dependent binding with fitted curves of CaM₃₄ (b) and CaM₁₂₃₄ (d) (0.1–50 μM) to CT1 at ≈ free, 100 nM, 10 μM and 2 mM Ca²⁺. Band densities were converted into molar ratio (CaM₃₄/GST-CT1) and plotted against total [CaM₃₄] or [CaM₁₂₃₄] with symbols as indicated in each graph and with mean ± SE (n = 3–10)

The bindings of CaM₃₄ and CaM₁₂₃₄ could be fitted well with the two-site model, and the parameters obtained (Table 1) revealed that not only K _d values but also B _max values had no obvious Ca²⁺-dependent change, suggesting that the C-lobe of CaM may be critical for the Ca²⁺-dependent binding of CaM to CT1.

Concentration- and Ca²⁺-dependent binding of CaM and its mutants to NT

Dick et al. [9] have reported that CaM could bridge the N- and C-terminal tails of Ca_V1.2 channels to control CDI, but Benmocha et al. [10] indicated that CaM did not foster an interaction between the CaM-binding peptides of N- and C-terminal tails. Thus, we examined the bindings of CaM and its mutants (CaM₁₂₃₄ and CaM₃₄) to NT, the GST-fusion peptide of Ca_V1.2 N-terminal tail, to investigate whether CaM is possible to bind the Ca_V1.2 N-terminal tail with its N-lobe. We found that CaM and the mutants did not show clear binding to NT at low [Ca²⁺] (≈free and 100 nM). When [Ca²⁺] was increased to 10 μM and 2 mM, CaM but not CaM₁₂₃₄ showed binding to NT (Fig. 4a), indicating that the binding between CaM and NT was Ca²⁺-dependent. The curves of CaM’s binding to NT were appropriately fitted with the one-site model (Fig. 4b), and the estimated B _max and K _d values are summarized in Table 2. CaM₃₄, in which Ca²⁺ binding is thought to be restricted to the N-lobe, also showed detectable binding to NT in the presence of Ca²⁺ (10 μM and 2 mM) (Fig. 4c, d; Table 2), implying that the N-lobe of CaM may play an important role in the binding of CaM to the N-terminal tail of Ca_V1.2.

Fig. 4

CaM and CaM₃₄ binding to the NT peptide of the Ca_V1.2 channel N-terminus. GST pull-down assay for CaM (a) or CaM₃₄ (c) binding to NT. Protein bands are shown as CBB staining at different [Ca²⁺]. CaM bands are indicated by arrows. Plots of concentration-dependent binding with fitted curves of CaM (b) or CaM₃₄ (d) (0.1–50 μM) to NT at free, 100 nM, 10 μM and 2 mM Ca²⁺. Band densities were converted into molar ratio (CaM/GST-NT) and plotted against total [CaM] or [CaM₃₄] with symbols as indicated in each graph and with mean ± SE (n = 3–10). *P < 0.05, **P < 0.01, compared with CaM or CaM₃₄ binding to NT at the corresponding CaM or CaM₃₄ concentration with Ca²⁺-free condition for (b) and (d), respectively

Table 2 Fitting parameters for the bindings of CaM and CaM₃₄ to NT

Discussion

In the present study, we have investigated the concentration- and Ca²⁺-dependencies of CaM binding with Ca_V1.2. It has been reported that more than one region in the C-terminal tail of Ca_V1.2 can bind CaM [3, 5, 8]. However, Xiong et al. [17] showed that, although synthetic short peptides could bind CaM, longer fragments could also bind to only one CaM molecule, suggesting that only one of these sequences binds CaM in the larger protein or that all the individual sequences contribute to a single CaM binding site. In the study of Mori et al. [13], L-type channels were fused to a single CaM molecule, and the results suggested that one CaM appears to be both necessary and sufficient to produce CDI for its associated channel. On the other hand, a crystallographic study suggested that more than one molecule of CaM can simultaneously bind to the C-terminal tail peptide of Ca_V1.2 [14]. Using a semi-quantitative pull-down method, Asmara et al. [15] also suggested that about two CaM molecules could bind to the longer C-terminal fragment of Ca_V1.2 comprising the preIQ and IQ regions.

Employing a similar pull-down assay method in the present study, we have found that the maximal binding of CaM with the C-terminal tail of Ca_V1.2, estimated as ${{\rm{B}_{\max_{1} }}+{\rm{B}_{\max_{2} }}}$, is over 1 mol/mol CaMs/peptide for all the four CaMs studied (Table 1). Since the C-terminal tail of Ca_V1.2 has more than one CaM binding site, and since CaM has two lobes, each of which can separately bind to the CaM binding sites of the binding partners [5, 18], it is reasonable to speculate that the Ca_V1.2 channel may have the capability to bind more than one CaM molecule. In particular, crystal experiments confirmed the existence of a complex constituted by two channels bound with four CaM molecules, although not all the lobes of CaMs are visible [14]. By considering the examined CaM and Ca²⁺ concentration ranges, it was observed that for wild-type CaM the binding ratio higher than 1 happens when [Ca²⁺] is higher than 10 μM and CaM concentration is higher than 1 μM (Fig. 2). Considering that the physiological concentrations of Ca²⁺ and CaM are less than 2 mM [2] and less than 10 μM [19], respectively, it is likely that, under physiological conditions from resting to earlier exciting, the channel may bind less or one CaM molecule. When Ca²⁺ and CaM concentrations increase further, such as when the cells are fully excited or Ca²⁺ is overloaded under pathophysiological condition, the C-terminal tail of the Ca_V1.2 channel may bind more than one CaM molecule. Additionally, provided that Ca_V1.2 channel in situ has a higher affinity than that observed with its fragment peptides in vitro, it may be possible that multiple molecules of CaM play a role in Ca²⁺-dependent regulation of the channel in physiological conditions.

It has been suggested that C- and N-lobes of CaM have distinct roles in the regulation of Ca²⁺ channels [6, 11, 12, 18]. It is reported that the C-lobe, when bound with Ca²⁺, triggers a rapid CDI process, whereas the N-lobe induces a distinct and relatively slow CDI in Ca_V1.2 channels [9]. Recently, a patch-clamp study showed that both N- and C-lobes of CaM are required for the Ca²⁺-dependent regulation of Ca_V1.2 channels [20, 21]. Thus, the role of each lobe of CaM remains to be elucidated. Although we demonstrated that all the three CaM mutants could bind with the CT1 peptide of Ca_V1.2 (Figs. 2, 3), the concentration-binding curve for CaM₁₂ shifted slightly upward with increasing [Ca²⁺], which is similar to that of wild-type CaM (Fig. 2). On the other hand, there was no significant shift for those of CaM₃₄ and CaM₁₂₃₄ (Fig. 3), suggesting that the C-lobe of CaM plays a dominant role in triggering the Ca²⁺-dependent regulation of Ca_V1.2.

Besides the C-terminal tail of the Ca_V1.2 channel, CaM has been reported to interact with the N-terminal tail [9, 15]. The functional significance of this region has been demonstrated for Ca_V1.2 in vivo by patch-clamp technique [9]. In this study, we have confirmed that CaM and CaM₃₄ could bind to NT at [Ca²⁺] ≥ 10 μM (Fig. 4). These results may imply that the Ca²⁺-bound N-lobe of CaM may play an important role in the binding of CaM to NT, and this idea is consistent with the hypothesis that CaM can bridge N- and C-terminal tails of Ca_V1.2 channel [9].

Based on the present and other data, we propose a hypothesis for the Ca²⁺/CaM-dependent regulation of Ca_V1.2 channels. At low and middle levels of [Ca²⁺], Ca_V1.2 channel may usually bind one CaM molecule, and its Ca²⁺-dependent regulation may be predominately conducted by the C-lobe of CaM. A Ca²⁺-dependent increase in the extent of CaM-bound channel may underlie CDF of the channel [20, 21]. When Ca²⁺ and CaM concentrations increased, the channel may bind with the second CaM molecule at its C terminal tail, and this may be related to CDI of the channel. However, we do not exclude the alternative possibility that the N-terminal tail of Ca_V1.2 binds with N-lobe while the C-terminal tail of the channel binds C-lobe of CaM (bridge formation), providing another conformation of CDI [9]. It is to be examined in future studies which conformation for CDI is occurring in the intact cells.