Calmodulin modulates the Ca²⁺-dependent inactivation and expression level of bovine Ca_V2.2 expressed in HEK293T cells

Abstract

Ca²⁺ influx through voltage-gated Ca²⁺ channels (Ca_Vs) at the plasma membrane is the major pathway responsible for the elevation of the intracellular Ca²⁺ concentration ([Ca²⁺]_i), which activates various physiological activities. Calmodulin (CaM) is known to be involved in the Ca²⁺-dependent inactivation (CDI) of several types of Ca_Vs; however, little is known about how CaM modulates Ca_V2.2. Here, we expressed Ca_V2.2 with CaM or CaM mutants with a Ca²⁺-binding deficiency in HEK293T cells and measured the currents to characterize the CDI. The results showed that Ca_V2.2 displayed a fast inactivation with Ca²⁺ but not Ba²⁺ as the charge carrier; when Ca_V2.2 was co-expressed with CaM mutants with a Ca²⁺-binding deficiency, the level of inactivation decreased. Using glutathione S-transferase-tagged CaM or CaM mutants as the bait, we found that CaM could interact with the intracellular C-terminal fragment of Ca_V2.2 in the presence or absence of Ca²⁺. However, CaM and its mutants could not interact with this fragment when mutations were generated in the conserved amino acid residues of the CaM-binding site. Ca_V2.2 with mutations in the CaM-binding site showed a greatly reduced current that could be rescued by CaM₁₂ (Ca²⁺-binding deficiency at the N-lobe) overexpression; in addition, CaM₁₂ enhanced the total expression level of Ca_V2.2, but the ratio of Ca_V2.2 present in the membrane to the total fraction remained unchanged. Together, our data suggest that CaM, with different Ca²⁺-binding abilities, modulates not only the inactivation of Ca_V2.2 but also its expression to regulate Ca²⁺-related physiological activities.

Highlights

• •CaM binds to the conserved IQ motif of Ca_V2.2 and triggers inactivation.
• •Mutations in the conserved IQ motif decrease the current density.
• •The IQ motif is the main CaM binding site.
• •CaM with Ca²⁺-binding deficiency at the N-lobe increases the Ca_V2.2 level.

1. Introduction

Intracellular Ca²⁺ concentration ([Ca²⁺]_i) changes are an important signal for a wide spectrum of cell activities from short-term neurotransmitter release to long-term control of gene expression (Berridge, 2014, Lian and Zheng, 2016, Mintz et al., 1995, Simms and Zamponi, 2014, Soderling and Derkach, 2000, Yagami et al., 2012). The voltage-gated Ca²⁺ channels (Ca_Vs) at the plasma membrane are encoded by various isoforms in different cells and are the main pathway for Ca²⁺ influx. One of the Ca_V isoforms, Ca_V2, is widely expressed in the central and peripheral nervous system (Coppola et al., 1994, Fujita et al., 1993, Mills et al., 1994, Westenbroek et al., 1992, Westenbroek et al., 1998) and is involved in synaptic transmission in most neurons (Dunlap et al., 1995, Olivera et al., 1994).

The opening of the Ca_Vs elevates the [Ca²⁺]_i to a μM level upon membrane depolarization from a resting level of approximately 50 nM (Simons, 1988). Both the Ca_V1s and Ca_V2s show Ca²⁺-dependent inactivation (CDI), and calmodulin (CaM), which has 4 EF-hand Ca²⁺ binding motifs, may play a role in this inactivation (Johny et al., 2013, Peterson et al., 1999, Soong et al., 2002, Tadross et al., 2008). CaM binds to the IQ motifs located at the C-terminals of several subtypes of voltage-gated Na⁺ and Ca²⁺ channels to regulate the channel activity in a Ca²⁺-dependent manner (Ben-Johny et al., 2015). CDI is a typical feedback inhibition for Ca²⁺ homeostasis (Eckert and Chad, 1984, Eckert and Tillotson, 1981, Tillotson, 1979, Zweifach and Lewis, 1995); in contrast, Ca²⁺ signaling also induces facilitation (Ca²⁺-dependent facilitation, CDF) in several types of Ca_Vs (Chaudhuri et al., 2007, Lee et al., 2000).

Ca_V2.2 shows CDI and voltage-dependent inactivation (VDI), but no CDF has been reported (Ben-Johny and Yue, 2014). Similar to other Ca_Vs, the CDI is mediated by CaM, but the detailed interaction is not clear. In this report, we verified that bovine Ca_V2.2 showed an apparent CDI, and CaM interacted with the intracellular C-terminal of Ca_V2.2 (Ca_V2.2-CT). Mutating the IF residues in the IQ motif to AF (Ca_V2.2^IF/AF) or AA (Ca_V2.2^IF/AA) not only decreased the current amplitude but also abolished the interaction with CaM. However, the co-expression of CaM₁₂ (Ca²⁺-binding deficiency at the N-lobe), which has a Ca²⁺-binding deficiency in the N-lobe, enhanced the currents of Ca_V2.2 and its mutants. Thus, these results reveal that the C- and N-terminals of CaM have differential effects on binding to Ca_V2.2 and regulating the channel activities.

2. Material and methods

2.1. Chemicals

Lipofectamine 2000, mouse monoclonal antibodies against the CaM and glutathione S-transferase (GST), Dulbecco's modified Eagle's medium, and other chemicals for cell culture were obtained from Invitrogen Inc. (Carlsbad, CA, USA). Mouse monoclonal antibodies against the Flag epitope and all additional chemicals, unless otherwise indicated, were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA).

2.2. Plasmid preparation

The plasmids for bovine Ca_V2.2 and their accessory subunits were generously provided by Dr. Aaron P. Fox (University of Chicago) (Currie and Fox, 2002). The protocol used to prepare rat brain cDNA and mutants of CaM was described previously using Pfu Ultra AD polymerase (Agilent Technologies, USA) (Lin et al., 2013, Shih et al., 2009). The primers for the AF mutant were (Forward) and (Reverse) ; the primers for the AA mutant were (Forward) and (Reverse) (shaded nucleotides indicate the mutated IF sequence). The triple-Flag-tagged Ca_V2.2s (T- Ca_V2.2s) were constructed by digesting a synthetic dsDNA (TCTAGACTTAAGACCGGTGCCACCATGGATTACAAGGATGACGACGATAAGGACTATAAGGACGATGATGACAAGGAC/TACAAAGATGATGACGATAAAG AATTCAAGCTTACCGGTATGGTCCGCTTCGGGGACGAGCTGGGCGGCCGCGGATCC, gene synthesized by Thermo Fisher Scientific, the underline indicates the triple-Flag sequence) with AflII and NotI sites and inserting the fragment at the 5′-terminal of Ca_V2.2. The clones with the intracellular C-terminal segment of T-Ca_V2.2s were digested with EcoRI from T-Ca_V2.2s (aa 1710-2332, gene number: NM_174632.2).

The primers for cloning the CaM from the rat brain were (Forward) (BamHI) and (Reverse) (XbaI) (shaded nucleotides indicated the restriction enzyme site); the construct was then subcloned into pcDNA3.1 plasmid. To synthesize the Ca²⁺-binding deficiency mutation, we mutated the last a.a., glutamate, of each EF-hand motif to glutamine using the following primers: EF1 Forward and Reverse ; EF2 Forward and Reverse ; EF3 Forward and Reverse ; EF4 Forward , and Reverse (shaded nucleotides indicate the mutated ones). All of the primers listed are in the 5′ to 3′ direction.

2.3. Transfection of HEK293T cells

For transient expression of the genes in HEK293T cells grown in a 35-mm dish, we mixed α_1B, β_2a, and α₂δ (1 μg total at a ratio of 1:1:1 and 0.1 μg of a green fluorescence protein (GFP) plasmid) with Lipofectamine 2000 according to the manufacturer's instructions. We used GFP fluorescence to identify transfected cells and performed experiments 24–36 h after transfection.

2.4. Protein extraction

HEK293T cells were dissolved in a lysis buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and 50 mM Tris, pH 7.5, Bioman Inc., Taiwan) containing a protease inhibitor cocktail (Set V, 1:100 dilution, Calbiochem, La Jolla, California, USA). The lysates were centrifuged at 1000 ×g for 30 min, and the supernatant was collected as the total lysate.

2.5. GST pulldown assays

We purified the GST-fused CaM, CaM₁₂, CaM₃₄ or CaM₁₂₃₄ (has no Ca²⁺-binding ability at both the N- and C-lobes) expressed in E. coli as described previously (Chou et al., 2015) and used a Bradford-based protein assay kit (Bio-Rad, USA) to determine the protein concentration. To pull down interacting proteins, we incubated the GST-fused protein or GST with GSH-Sepharose 4B beads (GF Healthcare, USA) following the protocol suggested by the manufacturer. We mixed the beads with cell lysate at room temperature for 1 h or 4 °C overnight in a lysis buffer. The proteins that bound to the beads were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blotting.

2.6. Electrophysiological recording

The recording procedure was described previously (Shih et al., 2009). The recordings were performed at room temperature with an EPC-10 amplifier and were controlled by the Pulse program (HEKA Elektronik, Lambrecht/Pfalz, Germany). In brief, a cell was incubated in NMG buffer (in mM, 130 NMG, 20 glucose, 10 HEPES, 1 MgCl₂·6H₂O, 2 KCl, 10 CaCl₂·2H₂O, pH 7.2 with KOH, 300–305 mOsm) and patched in whole-cell mode. For the Ba²⁺ current measurement, Ca²⁺ was replaced with 10 mM Ba²⁺. The membrane potential was held at −70 mV and depolarized to various potentials for channel activation. The pipette solution consisted of (in mM) 120 aspartic acid, 5 MgCl₂, 40 HEPES, 0.1 EGTA, 2 ATP, and 0.3 GTP, pH 7.3 with CsOH (310 mOsm/kg). The R₂₅₀ ratio was measured by holding at −70 mV and depolarizing for 250 ms from −50 to +120 mV. The activation curves were created from the tail currents obtained by a 10-ms depolarization to various potentials and used for the analysis of V_1/2 and slope.

2.7. Biotinylation

The HEK293T cells were seeded on poly-L-lysine-coated 35-mm dishes and transfected with different plasmids as previously described. The cells were incubated on ice and washed with D-PBS buffer supplemented with 0.5 mM CaCl₂ and 2 mM MgCl₂, followed by 1 mg/mL sulfo-NHS-LC-biotin (Thermo Fisher Scientific) in 1 mL of D-PBS on ice with gentle rocking for 1 h. After being washed with 100 mM of glycine in PBS (in mM, 142 NaCl, 2 KCl, 8 Na₂HPO₄·7H₂O, and 1.5 NaH₂PO₄·H₂O, pH 7.2) two times and TBS (in mM, 20 Tris-HCl, and 150 NaCl, pH 7.4) one time, solubilization was performed using 400 μL of lysis buffer (in mM, 150 NaCl, 5 EDTA, and 50 Tris-HCl, pH 7.6 and 1% Triton-X 100 with dithiothreitol, 1 phenylmethylsulfonyl fluoride and protease inhibitors). The cell lysates were incubated with 30 μL streptavidin-agarose beads (Thermo Scientific) overnight at 4 °C. The beads were washed once in lysis buffer, followed by another wash in high salt buffer (in mM, 500 NaCl, 5 EDTA, and 50 Tris-HCl, pH 7.6 and 0.1% Triton-X 100) and low salt buffer (in mM, 2 EDTA, and 10 Tris-HCl, pH 7.6 and 0.1% Triton-X 100). Then, Laemmli sample buffer (containing 6% SDS) was added to the samples, and the samples were boiled at 70 °C for 10 min and prepared for Western blot analysis.

2.8. Statistical analysis

Data are presented as the mean ± SEM from at least three different batches of cells and were analyzed by one-way ANOVA with Fisher's post hoc test. Differences were considered significant when the p value was less than 0.05.

3. Results

3.1. Bovine Ca_V2.2 shows Ca²⁺-dependent inactivation

To examine the inactivation of bovine Ca_V2.2 currents, we expressed bovine α_1B and accessory subunits in HEK293T cells and measured the whole-cell currents in buffers containing Ca²⁺ or Ba²⁺ as the charge carrier (Fig. 1). When depolarized to various potentials for 250 ms from a holding potential of −70 mV, the representative current traces of cells incubated in buffer containing Ca²⁺ but not Ba²⁺ showed a fast inactivation after reaching the maxima (Fig. 1A). The current-voltage relationship showed that the depolarization potential for the maximum peak inward current was +30 mV for Ca_V2.2 using either Ca²⁺ or Ba²⁺ as the charge carrier (Fig. 1B). The residual currents at the end of the depolarization (R₂₅₀) of a representative cell were 27.8 or 77.5% of the peak current using Ca²⁺ (R₂₅₀/Ca²⁺) or Ba²⁺ (R₂₅₀/Ba²⁺), respectively, as the charge carrier at a depolarization potential of +30 mV (Fig. 1C). Fig. 1D shows that the average R₂₅₀/Ca²⁺ and R₂₅₀/Ba²⁺ were 17.4 ± 2.3 (n = 31) and 79.0 ± 2.5% (n = 15), respectively. These results demonstrate that Ca²⁺ is responsible for the majority of the inactivation of bovine Ca_V2.2 currents.

Fig. 1.

Bovine Ca_V2.2 shows Ca²⁺-dependent inactivation. HEK293T cells transfected with Ca_V2.2 were whole-cell patched in voltage-clamp mode and depolarized to various potentials from a holding potential of −70 mV for 250 ms. A. The current traces of a representative cell at different potentials using Ca²⁺ (middle panel) or Ba²⁺ (bottom panel) as the charge carrier. The black line indicates the trace obtained at +30 mV. B. The normalized current-voltage relationship. The peak inward current measured using Ca²⁺ (n = 45) or Ba²⁺ (n = 9) as the charge carrier at different potentials was normalized to the respective maxima of each single cell and then averaged. C. The current evoked by a step depolarization from −70 to +30 mV for 250 ms was recorded using Ca²⁺ (red) or Ba²⁺ (blue) as the charge carrier. The residual current at the end of the depolarization was normalized to the peak current as the R₂₅₀/Ca²⁺ or R₂₅₀/Ba²⁺. D. The average R₂₅₀/Ca²⁺ and R₂₅₀/Ba²⁺ of Ca_V2.2. Data are presented as the mean ± SEM. **, p < 0.01 compared to the Ca²⁺ group using Student's t-test.

3.2. Both lobes of CaM modulate the CDI effect

CaM has 4 Ca²⁺ binding sites with 2 at each of the N- or C-lobes, and each lobe has different effects in modulating various Ca_Vs (Ben-Johny et al., 2015, Yang et al., 2006). To verify the effect of each lobe of CaM in regulating Ca_V2.2, we measured the Ca_V2.2 currents using Ca²⁺ or Ba²⁺ as the charge carrier with co-expression of CaM or Ca²⁺-binding-deficient mutants in HEK293T cells (Fig. 2 & Table 1). The results indicated that the current traces in cells expressing CaM showed an apparent inactivation when using Ca²⁺ as the charge carrier; in contrast, the inactivation level was decreased in cells expressing CaM₁₂, CaM₃₄, and CaM₁₂₃₄ with Ca²⁺-binding deficiencies at the N-lobe, C-lobe, and both, respectively. The results (Fig. 2D & Table 1) showed that CaM mutants, CaM₁₂, CaM₃₄, and CaM₁₂₃₄, all showed a significant increase in R₂₅₀/Ca²⁺. In contrast, CaM and CaM₁₂ significantly lowered the R₂₅₀/Ba²⁺; CaM₃₄ and CaM₁₂₃₄ had no effect on the R₂₅₀/Ba²⁺. However, CaM and CaM₁₂ significantly decreased and increased, respectively, the peak inward current at +30 mV (I₃₀/Ca²⁺) when using Ca²⁺ as the charge carrier; CaM₃₄ and CaM₁₂₃₄ had no effect when compared with that of the Ca_V2.2 alone. However, using Ba²⁺ as the charge carrier enhanced the current density (I₃₀/Ba²⁺) except in the co-expression of CaM₁₂ (Table SI). These results suggest that the N- and C-lobes of CaM have differential effects in modulating the current amplitude and inactivation.

Fig. 2.

Ca²⁺-binding-deficient CaM mutants increase the R₂₅₀/Ca²⁺. We co-expressed Ca_V2.2 with CaM and CaM mutants with binding deficiencies at the N-, C-, and both lobes (CaM₁₂, CaM₃₄, and CaM₁₂₃₄, respectively) in HEK293T cells. The currents were recorded at +30 mV using Ca²⁺ or Ba²⁺ as the charge carrier, and the R₂₅₀/Ca²⁺ and R₂₅₀/Ba²⁺ were analyzed. A. Schematic representation of the CaM and CaM mutants. B. Representative normalized current traces (each trace was normalized to the inward peak current) from cells expressing CaM or mutants. C. Average R₂₅₀/Ca²⁺ and R₂₅₀/Ba²⁺. The numbers on each column refer to the number of cells used in each group. Data are presented as the mean ± SEM and were analyzed by one-way ANOVA with Fisher's post hoc test. *: p < 0.05, **: p < 0.01, and ***: p < 0.001 compared to the same expression group using Ca²⁺ as the charge carrier. #: p < 0.05 and ##: p < 0.001 compared to the R₂₅₀/Ba²⁺ of the CaM group. D. Plot of R₂₅₀/Ca²⁺ against I₃₀/Ca²⁺. The R₂₅₀/Ca²⁺ of each cell in Fig. 2C was plotted against the I₃₀/Ca²⁺. The lines indicated the linear regression of each group.

Table 1.

The basic properties of Ca_V2.2 with mutations in the IQ motif.

Co-expression		V1/2 (mV)	Slope (mV)	I30/Ca2+ (pA/pF)	R250/Ca2+ (%)	R250/Ba2+ (%)
CaV2.2	Mock	18.7 ± 2.7 n = 20	6.2 ± 0.8 n = 20	−71.1 ± 8.4 n = 37	17.4 ± 2.3 n = 31	79.0 ± 2.5 n = 13
	CaM	13.5 ± 2.7 n = 8	4.6 ± 1.0 n = 8	−29.9 ± 11.7* n = 10	23.8 ± 5.1 n = 11	60.7 ± 5.2* n = 10
	CaM12	24.7 ± 2.4 n = 4	7.1 ± 0.6 n = 4	−170.3 ± 25.7** n = 6	45.1 ± 4.2* n = 11	68.7 ± 4.3* n = 8
	CaM34	25.5 ± 3.2 n = 5	7.7 ± 1.3 n = 5	−78.0 ± 16.5 n = 8	45.0 ± 4.3* n = 13	82.5 ± 3.8 n = 6
	CaM1234	21.9 ± 2.9 n = 16	5.9 ± 0.8 n = 16	−51.8 ± 11.3* n = 11	41.0 ± 4.1* n = 17	82.1 ± 4.8 n = 5
CaV2.2IF/AF	Mock	31.6 ± 1.9 n = 7	8.8 ± 1.2 n = 7	−38.5 ± 9.4 n = 17	21.5 ± 4.0 n = 17	68.5 ± 3.9 n = 4
	CaM	26.2 ± 3.4 n = 5	8.6 ± 1.9 n = 5	−52.8 ± 15.0 n = 9	24.9 ± 3.5 n = 8	64.7 ± 6.0 n = 11
	CaM12	15.4 ± 5.7 n = 5	6.8 ± 1.6 n = 5	−109.5 ± 40.9 n = 7	45.2 ± 2.8 n = 8	61.5 ± 5.5 n = 5
	CaM34	37.7 ± 2.5 n = 5	14.2 ± 3.2 n = 5	−23.7 ± 7.9 n = 12	25.0 ± 4.1 n = 12	67.6 ± 5.5 n = 6
	CaM1234	30.7 ± 2.8 n = 5	9.1 ± 0.6 n = 5	−33.3 ± 9.3 n = 16	33.5 ± 7.6 n = 7	69.5 ± 8.6 n = 5
CaV2.2IF/AA	Mock	41.4 ± 2.0 n = 6	13.2 ± 1.1 n = 6	−13.4 ± 3.3 n = 9	19.5 ± 2.6 n = 6	57.7 ± 4.8 n = 6
	CaM	30.0 ± 4.9 n = 6	7.9 ± 2.1 n = 6	−60.5 ± 19.9 n = 9	25.2 ± 5.7 n = 7	63.4 ± 6.8 n = 4
	CaM12	10.6 ± 13.1 n = 3	6.4 ± 2.1 n = 3	−118.8 ± 41.3 n = 4	37.8 ± 5.6 n = 6	82.6 ± 4.3 n = 4
	CaM34	7.6 ± 10.1 n = 3	0.8 ± 0.5 n = 3	−63.1 ± 20.3 n = 7	22.5 ± 2.6 n = 5	63.4 ± 6.8 n = 3
	CaM1234	25.0 ± 6.4 n = 4	8.1 ± 2.5 n = 4	−20.0 ± 4.4 n = 9	16.0 ± 8.1 n = 8	42.7 ± 3.1 n = 3

Data are presented as the mean ± SEM and were analyzed by one-way ANOVA with Fisher’s post hoc test. *: p < 0.05 and **: p < 0.01 when compared to the Mock group.

3.3. Mutating the IF residues reduces the current amplitude

The IQ motif of Ca_V1s and Ca_V2.1 is responsible for CaM binding and the CDI effect (Ben-Johny et al., 2015). To assess the importance of this motif of Ca_V2.2, we expressed Ca_V2.2^IF/AF or Ca_V2.2^IF/AA in HEK293T cells and measured the currents after depolarization to different potentials for 250 ms (Fig. 3 and Table 1). The normalized current-voltage relationship (Fig. 3B) showed that the voltages of the maximum inward current were positively shifted in these mutants. The activation curves (Fig. 3D and Table 1) showed that the V_1/2 values were significantly shifted from 18.7 ± 2.7 mV (n = 20) with wild-type Ca_V2.2 to 31.6 ± 1.9 mV (n = 7, p < 0.01) and 41.4 ± 2.0 mV (n = 6, p < 0.01) with Ca_V2.2^IF/AF and Ca_V2.2^IF/AA, respectively. The I₃₀/Ca²⁺ of Ca_V2.2^IF/AF and Ca_V2.2^IF/AA was significantly reduced to −38.5 ± 9.4 pA/pF (n = 17, p < 0.05) and −13.4 ± 3.3 pA/pF (n = 9, p < 0.001), respectively, from that of the wild-type, with a value of −71.1 ± 8.4 pA/pF (n = 37). Therefore, the conserved IF residues are important in maintaining the current amplitude and current-voltage relationship.

Fig. 3.

Mutations in the IF residues of Ca_V2.2 reduce the current. Ca_V2.2 with mutations in the IF residues (Ca_V2.2^IF/AF and Ca_V2.2^IF/AA) was expressed in the HEK293T cells. A. Schematic representation of Ca_V2.2 and the constructs. B. The current-voltage relationship. The cell was whole-cell patched and depolarized for 250 ms from a holding potential of −70 mV to various potentials using Ca²⁺ as the charge carrier. The peak current recorded at each potential was normalized to the inward maxima of each cell. C. The activation curves. Cells were depolarized for 10 ms from a holding potential of −70 mV to various potentials, and the tail inward current obtained at each potential was normalized to the maxima. D. The I₃₀/Ca²⁺ of Ca_V2.2 and the mutants. The current was evoked by a depolarization to +30 mV for 250 ms, and the peak inward current was measured. Data are presented as the mean ± SEM and were analyzed by one-way ANOVA with Fisher's post hoc test. *: p < 0.05 and ***: p < 0.001 compared to the wild-type.

3.4. The IF amino acid residues are important for current inactivation

To further assess the importance of the conserved IF residues of Ca_V2.2 in current inactivation, we co-expressed Ca_V2.2^IF/AF or Ca_V2.2^IF/AA with CaM and measured the currents (Fig. 4 and Table 1). Representative traces (Fig. 4A and B) of Ca_V2.2^IF/AF and Ca_V2.2^IF/AA showed a higher level of inactivation when using Ca²⁺ as the charge carrier than Ba²⁺. Fig. 4C and D shows that the average R₂₅₀/Ca²⁺ of Ca_V2.2^IF/AF and Ca_V2.2^IF/AA was similar to that of Ca_V2.2; however, the R₂₅₀/Ba²⁺ was 68.5 ± 3.9% (n = 4, p < 0.05) and 57.7 ± 4.8% (n = 6, p < 0.01), respectively, which were both significantly lower than that of the wild-type form (79.0 ± 2.5%, n = 13). CaM₁₂ significantly increased the R₂₅₀/Ca²⁺ of Ca_V2.2^IF/AF to 45.2 ± 2.8% (n = 8, p < 0.05) compared to that of the CaM group, but CaM₃₄ and CaM₁₂₃₄ had little effect on the R₂₅₀/Ca²⁺. In contrast, the R₂₅₀/Ba²⁺ values for Ca_V2.2^IF/AF with different CaM constructs were approximately the same in a range between 60 and 70%. For Ca_V2.2^IF/AA, the co-expression of CaM and mutants did not have any significant effect on the R₂₅₀/Ca²⁺ and R₂₅₀/Ba²⁺. The activation curves of the wild-type and mutated channels co-expressing CaM and CaM mutants are shown in Fig. S1; the calculated V_1/2 and slope are listed in Table 1. These results suggest that mutations in the IF residues blocked the effects of the CaM mutants, except CaM₁₂, in suppressing the current inactivation.

Fig. 4.

CaM₁₂ increases the R₂₅₀/Ca²⁺ of mutated Ca_V2.2. The conserved IF residues of Ca_V2.2 were mutated to AF (Ca_V2.2^IF/AF) or AA (Ca_V2.2^IF/AF) and expressed in HEK293T cells. A & B. The representative current traces from cells expressing Ca_V2.2^IF/AF and Ca_V2.2^IF/AA co-expressing CaM and CaM mutants. The patched cell was depolarized to +30 mV from a holding potential of −70 mV for 250 ms using Ca²⁺ (red traces) or Ba²⁺ (blue traces) as the charge carrier. C & D. Average R₂₅₀ of Ca_V2.2^IF/AF and Ca_V2.2^IF/AF, respectively. Data are presented as the mean ± SEM and were analyzed by one-way ANOVA with Fisher's post hoc test. *: p < 0.05 and ***: p < 0.001 compared to the same expression group or as indicated.

3.5. CaM₁₂ enhances the attenuated current density

In addition to the effects on current inactivation, the mutations in the IF residues greatly reduced the current amplitude (Table 1). Compared to the group expressing Ca_V2.2^IF/AF and CaM, which had an I₃₀/Ca²⁺ of −52.8 ± 15.0 pA/pF (n = 9), CaM₁₂ substantially enhanced the I₃₀/Ca²⁺ to −109.5 ± 40.9 pA/pF (n = 7, p < 0.05); in contrast, CaM₃₄ and CaM₁₂₃₄ slightly decreased the I₃₀/Ca²⁺ but not significantly. For Ca_V2.2^IF/AA, CaM, CaM₁₂, and CaM₃₄ could slightly but not significantly enhance the I₃₀/Ca²⁺. By plotting the R₂₅₀/Ca²⁺ against the I₃₀/Ca²⁺ (Fig. 2D) from cells expressing Ca_V2.2 and CaM/mutants, the linear regression curves showed an inverse relationship, except for the CaM₁₂ group, which had a R₂₅₀/Ca²⁺ above 40% regardless of the current density. In addition, the I₃₀/Ba²⁺ were larger than those I₃₀/Ca²⁺ of the same groups except the CaM₁₂ group (Table SI). For Ca_V2.2 wild type and mutants, the I₃₀/Ca²⁺ and I₃₀/Ba²⁺ were about the same in the co-expression of CaM₁₂. These results indicated that the conserved IF residues and each lobe of CaM are important in regulating the current density.

3.6. Differential binding of CaM mutants to the C-terminal fragment of Ca_V2.2

To confirm that CaM interacts with Ca_V2.2 at the C-terminal, we used GST-tagged CaM or Ca²⁺-binding-deficient mutants as the bait to pull down Ca_V2.2-CT expressed in HEK293T cells (Fig. 5A–D). A representative Western blot shows that GST-CaM pulled down a protein with a MW similar to the expected size of Ca_V2.2-CT (∼72 kD) in the presence of Ca²⁺ (100 μM), EGTA (5 μM), or no addition; however, GST-CaM₁₂ and -CaM₁₂₃₄ could only interact with this fragment in the presence of Ca²⁺ but with a lower capacity than that of CaM. CaM₃₄ did not show any interaction with the Ca_V2.2-CT. We then mutated the conserved IF residues to AA (Ca_V2.2-CT^IF/AA) for pulldown assays to characterize the importance of the IF residues in this interaction. Western blot analysis (Fig. 5E) showed that little Ca_V2.2-CT^IF/AA was observed regardless of the presence of CaM or mutants with or without Ca²⁺. These results illustrate the importance of the IF residues in the binding of CaM to Ca_V2.2 and indicate that each lobe of CaM has differential contributions to this interaction.

Fig. 5.

CaM interacts with the C-tail fragment of Ca_V2.2. The lysates from HEK293T cells expressing the intracellular C-tail fragment of the wild-type (CT, 72 kDa) (A–D) or mutant Ca_V2.2 (CT^IF/AA) (E) were used for pulldown assays with GST-tagged CaM and mutants (43 kDa) as the baits. The reaction buffer for the assays had no extra addition (None) or contained 100 μM of Ca²⁺ (Ca) or 5 μM of EGTA (EGTA). The pulldown fraction was then analyzed by Western blotting using a monoclonal antibody against the C-terminal of Ca_V2.2 (CT) or CaM.

3.7. CaM₁₂ enhances the expression level of Ca_V2.2

Because CaM and the mutants affected the I₃₀/Ca²⁺, we examined the total expression level and the fraction of Ca_V2.2 localized at the cell membrane (Fig. 6). The lysates from cells transfected with Ca_V2.2 showed a protein band with a MW greater than 250 kD that was not observed in the control group without Ca_V2.2 expression (Fig. 6A). With co-expression of CaM and the mutants, the lysate containing CaM₁₂ showed an increased band intensity compared to that of the other groups. The average total Ca_V2.2 level from cells co-expressing CaM₁₂ was 1.6 ± 0.1 (n = 3, p < 0.05) when normalized to the group with CaM overexpression, which was significantly higher than that with CaM overexpression (Fig. 6B). CaM₃₄ (1.1 ± 0.3) resulted in a similar level of Ca_V2.2 expression to that of the CaM group; CaM₁₂₃₄ (0.7 ± 0.3, n = 3, p = 0.06) slightly reduced the expression level but not significantly. Because the number of channels at the plasma membrane determines the current density, we used biotinylation to label the proteins at the plasma membrane and analyzed the ratio of Ca_V2.2 at the membrane to the total lysate (Fig. 6C–D). Western blot analysis showed that the antibody against Ca_V2.2 could recognize a protein in the biotinylation fraction with a MW similar to that in the total lysate. After analyzing the intensities of these bands, the Ca_V2.2 located at the membrane surface was 40–50% of the total Ca_V2.2 in all cell groups. These results indicate that CaM₁₂ increases the total amount of Ca_V2.2 in the cells and plasma membrane.

Fig. 6.

CaM₁₂ enhances the total expression level of Ca_V2.2. A. Representative Western blot of Ca_V2.2 in total lysates. Cell lysates isolated from HEK293T cells with no transfection (None), transfected with T-Ca_V2.2 only (T-Ca_V2.2, Mock), or transfected with co-expression of CaM, CaM₁₂, CaM₃₄, and CaM₁₂₃₄ were used for Western blot analysis using antibodies against FLAG and β-actin. B. Normalized total Ca_V2.2 level. The level of T-Ca_V2.2 was normalized to the level of β-actin in each sample; the values were then normalized to the value of the CaM group in each independent experiment. C. Representative staining of Ca_V2.2 after being biotinylated. Membrane proteins were first labeled with biotin and then isolated with avidin. The T-Ca_V2.2 in the total lysate (Total) and avidin-purified fraction (Biotinylated) was analyzed by an antibody against FLAG. D. The ratio of Ca_V2.2 in the membrane fraction to the total lysate. Data are presented as the mean ± SEM from at least 3 independent experiments and were analyzed by one-way ANOVA with Fisher's post hoc test. *: p < 0.05 compared to the CaM group.

4. Discussion

Ca²⁺ influx through the Ca_Vs is the major elicitor of exocytosis and many other Ca²⁺-related activities in excitable cells; thus, regulating the kinetics of Ca_Vs could be an effective way to modulate different cellular functions (Dubel et al., 1992, Williams et al., 1992). Our results demonstrated that CaM binds to the conserved IQ motif of Ca_V2.2 at the intracellular C-terminal to trigger CDI. In addition, CaM₁₂ enhanced the current amplitude and the total expression level of Ca_V2.2 at the cell membrane. Therefore, the N- and C-lobes of CaM have differential effects in regulating Ca_V2.2 activity.

For Ca_V1s, Ca_V2.1, and Ca_V2.3, binding of CaM to the C-terminal IQ motif determines the CDI (DeMaria et al., 2001, Yang et al., 2006). Several reports utilizing a gene-shuffled chimeric C-terminal of Ca_V2.2 and Ca_V1s have also suggested that CaM binds to the C-terminal end and promotes CDI of the chimeric channels (Kim et al., 2008, Mori et al., 2008). The IQ segments of Ca_V2.1-3 interact with CaM as determined by X-ray crystallography and isothermal titration calorimetry (Fallon et al., 2009, Kim et al., 2008, Wang et al., 2014). The X-ray structure analysis shows that the IQ-helix peptides of Ca_V1.2 and Ca_V2s bind the pocket of CaM in the opposite orientation. The N-lobe of CaM interacts more with the a. a. residues located at the N-terminal portion of the Ca_V1.2 IQ peptide relative to the conserved IQ residues but the C-terminal portion of the IQ peptide of the Ca_V2s and vice versa. Even so, the CDI is determined by the interaction between the CaM N-lobe and the C-terminal portion of the IQ motif; therefore, mutations at the N-lobe (CaM₁₂ and CaM₁₂₃₄) lose the ability to induce CDI (Liang et al., 2003). However, our results showed that losing the Ca²⁺-binding ability at either the N- or C-lobes of CaM decreases the CDI to a similar level, suggesting that either interaction is necessary and sufficient to maintain the CDI. We mutated the last a. a., E, in each EF-hand motif of CaM to Q to reduce the Ca²⁺-binding capability, while Liang et al. (2003) converted the first residues, D, of each motif to A. This discrepancy may differentially affect the functions of CaM and explain the difference of the results.

An amino-terminal Ca²⁺/CaM binding segment (NSCaTE, N-terminal Spatial Ca²⁺ Transforming Element) of Ca_V1.3 is known to interact with the N-lobe of CaM (Liu and Vogel, 2012). In the N-portion of the C-terminus of Ca_V1s and Ca_V2s, the pre-IQ domain and IQ domain are also CaM binding sites (Ben-Johny et al., 2015, Johny et al., 2013). The NSCaTE element can interact with Ca²⁺/CaM prebound to an IQ domain peptide, suggesting the possible bridging of the channel amino- and carboxyl-termini; however, Ca_V2.2 does not contain the NSCaTE element (Taiakina et al., 2013). Therefore, CaM may interact with the Ca_V2.2 mostly through the intracellular C-terminal.

Here, using pulldown assays and current measurement, we further demonstrated that the binding of CaM to Ca_V2.2 is not Ca²⁺-dependent, suggesting that CaM could interact with Ca_V2.2 at rest; in addition, the over-expression of CaM did not enhance the CDI, indicating that most Ca_V2.2 molecules are bound with CaM at rest, and a global elevation of [Ca²⁺]_i would activate the CDI (Ben-Johny and Yue, 2014, Dick et al., 2008, Dick et al., 2016, Few et al., 2012, Liang et al., 2003). The plot of the R₂₅₀/Ca²⁺ against the I₃₀/Ca²⁺ from cells expressing Ca_V2.2 shows an inverse relationship, supporting the global Ca²⁺ effect on current inactivation. The expression of CaM and the mutants showed an inverse relationship as well, except for CaM₁₂, which not only enhanced the I₃₀/Ca²⁺ but also maintained the R₂₅₀/Ca²⁺ above 40%. However, using Ba²⁺ as the charge carrier did not enhance the current amplitude in the co-expression of CaM₁₂. It is possible that CaM₁₂ blocks the CDI even when the current density is high or increases the VDI portion of the evoked current. The mechanism by which CaM₁₂ modulates the current inactivation needs to be investigated further.

In contrast to the wild-type CaM, the binding of CaM₁₂ to Ca_V2.2-CT requires Ca²⁺; therefore, when [Ca²⁺]_i is elevated, CaM₁₂ competes with the endogenous bound CaM and increases the R₂₅₀/Ca²⁺. Although CaM₃₄ is incapable of binding to Ca_V2.2-CT, CaM₃₄ may interfere with the interaction between endogenous CaM and Ca_V2.2 to increase the R₂₅₀/Ca²⁺. The Ca_V2.2-CT we constructed includes a Ca²⁺-binding motif (Delcour et al., 1993, Johny et al., 2013, Yang et al., 2006); upon Ca²⁺ binding, this motif may prepare Ca_V2.2-CT for CaM binding. As CaM can have different conformations when free or bound with Ca²⁺ (Fallon et al., 2009, Hultschig et al., 2004), CaM with a mutated N-lobe (CaM₁₂ and CaM₁₂₃₄) may have a conformation that allows Ca²⁺-dependent binding to Ca_V2.2.

The binding of CaM to the IQ motif of Ca_V2.2 is important in maintaining the current amplitude as mutations in the conserved IF residues greatly reduced the I₃₀/Ca²⁺, and the overexpression of CaM, especially CaM₁₂, rescued the current amplitude of these mutants. The pulldown assays showed that CaM and the mutants did not bind to Ca_V2.2-CT^IF/AA, and CaM₁₂ overexpression increased the total amount of Ca_V2.2 expressed in HEK293T cells but did not affect the fraction of Ca_V2.2 at the plasma membrane (Fig. 6). This may partly explain how CaM₁₂ overexpression enhances the current. However, it is not clear how CaM₁₂ modulates the R₂₅₀/Ca²⁺ level. It is possible that most of the current either enhanced by CaM₁₂ or decreased by the channel mutations was due to CDI, but this needs to be further characterized.

The β and α₂δ subunits of Ca_Vs are involved in modulating the kinetics and amplitude of the currents, as well as targeting the channels to the plasma membrane (Brice and Durward, 1997, Singer et al., 1991). The transient over-expression of α₂δ-1, α₂δ-2 and α₂δ-3 subunits in cultured hippocampal neurons increases not only the presynaptic abundance of P/Q-type channels but the probability of vesicular release in response to a single action potential (Hoppa et al., 2012). The cytosolic β subunits have a chaperone-like effect in promoting the functional expression of the subunits of Ca_V2s at the plasma membrane (Bichet et al., 2000, Brice and Durward, 1997, Raghib et al., 2001). In addition, β subunits control the gating properties of the Ca_Vs and hyperpolarize the voltage-dependence of activation, as well as increase the maximum open probability resulting in increasing the macroscopic current density (Matsuyama et al., 1999, Meir et al., 2000, Neely et al., 1993). Our results showed that CaM₁₂ could enhance the total expression level of Ca_V2.2 but not the ratio of the membrane localization. It is not clear how CaM regulates the membrane targeting of Ca_V2.2 and needs to be further characterized.

5. Conclusion

Our findings suggest that CaM binds to Ca_V2.2 via the conserved IQ motif to modulate the expression level and current density at rest or low [Ca²⁺]_i; when [Ca²⁺]_i elevates, the bound CaM enhances CDI. Because CaM₁₂ and CaM₃₄ show different effects in modulating these above-mentioned activities, the N- and C-lobes of CaM work differentially in modulating Ca_V2.2. In addition to providing various binding sites for activity regulation, the C-terminus of Ca_V1.2 and Ca_V2.1 could be cleaved by a protease, such as calpain, and the dissociated distal fragments may either interact with the proximal C-terminus to inhibit the channel activity or translocate to the nucleus, acting as a transcription factor (Abele and Yang, 2012, Hell et al., 1996). These results reveal that the versatile pathways for regulating channel activity via the C-terminus of Ca_Vs and CaM, with different Ca²⁺-binding abilities, are an immediate regulatory factor for various cellular functions (Ben-Johny et al., 2015).

Appendix A. Supplementary data

The following is the supplementary data related to this article: