Zn²⁺ Sensitivity of High- and Low-Voltage Activated Calcium Channels

Abstract

The essential cation zinc (Zn²⁺) blocks voltage-dependent calcium channels in several cell types, which exhibit different sensitivities to Zn²⁺. The specificity of the Zn²⁺ effect on voltage-dependent calcium channel subtypes has not been systematically investigated. In this study, we used a transient protein expression system to determine the Zn²⁺ effect on low- and high-voltage activated channels. We found that in Ba²⁺, the IC₅₀ value of Zn²⁺ was α₁-subunit-dependent with lowest value for Ca_V1.2, and highest for Ca_V3.1; the sensitivity of the channels to Zn²⁺ was approximately ranked as Ca_V1.2 > Ca_V3.2 > Ca_V2.3 > Ca_V2.2 = Ca_V 2.1 ≥ Ca_V3.3 = Ca_V3.1. Although the Ca_V2.2 and Ca_V3.1 channels had similar IC₅₀ for Zn²⁺ in Ba²⁺, the Ca_V2.2, but not Ca_V3.1 channels, had ∼10-fold higher IC₅₀ to Zn²⁺ in Ca²⁺. The reduced sensitivity of Ca_V2.2 channels to Zn²⁺ in Ca²⁺ was partially reversed by disrupting a putative EF-hand motif located external to the selectivity filter EEEE locus. Thus, our findings support the notion that the Zn²⁺ block, mediated by multiple mechanisms, may depend on conformational changes surrounding the α₁ pore regions. These findings provide fundamental insights into the mechanism underlying the inhibitory effect of zinc on various Ca²⁺ channel subtypes.

INTRODUCTION

The regulation of calcium entry into cells via voltage-dependent calcium channels (VDCCs) plays a fundamental role in controlling synaptic transmission, membrane excitability, muscle contraction, rhythmic activity, gene transcription, and signal transduction pathways (1,2). Therefore, elucidating the molecular mechanisms regulating calcium channel conductivity is essential for a greater comprehension of cell biology.

Zinc, an essential transition divalent cation, is involved in maintaining and regulating cellular and subcellular functions of virtually all cells. Four major roles of Zn²⁺ have been reported: 1), it binds tightly to metalloenzymes and serves as a cofactor in gene expression and enzymatic reactions (3); 2), it is abundant in the brain, where it is localized in the presynaptic terminals, is coreleased with glutamate (4,5), and regulates neuronal excitability and synaptic plasticity (6,7); 3), it is involved in many pathological neuromodulatory events, and the accumulation of Zn²⁺ during and after transient global ischemia (8,9) can also affect cardiac functions (10,11); and 4), it regulates conductivity of VDCCs (12–14) and affects Ca²⁺ signaling.

Zn²⁺ block of VDCCs has been found in various neuronal preparations from different species, including DRG neurons (12,15), hypothalamic neurons (13,16), paleocortical neurons (17), thalamic relay neurons (14), and pelvic neurons (18). The sensitivity of native calcium channels to Zn²⁺ is highly variable, with IC₅₀ values ranging from 7 μM (13) to 300 μM (15). In these studies, Zn²⁺ inhibition of the VDCCs was primarily affected by the animal species used and the cell type studied (14,19). Most neurons express multiple calcium channel subtypes, and each channel has a distinct physiological role. Thus, the sensitivity of specific calcium channels to zinc inhibition, which has not yet been determined, is critically important for a better understanding of their particular role in calcium conductance.

Inhibition of different calcium channel subtypes by various inorganic and organic calcium channel blockers (20,21) has been studied in cells in which the channel subtypes were transiently expressed. High-voltage–activated (HVA) calcium channels are heteromultimers consisting of α₁, α₂, β-, and δ-subunits. The α₁ subunit, which contains the channel pore region, encompassing the selectivity filter locus, plays the predominant role in determining channel conductance. Currently, 11 VDCC α₁ subunits have been cloned, including four HVA Ca_V1 L-type channels (Ca_V1.1/α_1S, Ca_V1.2/α_1C, Ca_V1.3/α_1D, and Ca_V1.4/α_1F); three HVA Ca_V2 non-L-type calcium channels (Ca_V2.1/α_1A, Ca_V2.2/α_1B, and Ca_V2.3/α_1E); and three low-voltage-activated (LVA) Ca_V3 T-type channels (Ca_V3.1/α_1G, Ca_V3.2/α_1H, and Ca_V3.3/α_1I) (22–24). The electrophysiological and pharmacological properties of the α₁ subunits have been well described. For instance, α_1C (Ca_V1.2) is found in dihydropyridine-sensitive L-type channels; α_1A (Ca_V2.1) in ω-agatoxin IVA-sensitive P-/Q-type channels; α_1B (Ca_V2.2) in ω-conotoxin GVIA-sensitive N-type channels; and α_1E (Ca_V2.3) in R-type channels. T-type calcium channels contain one of three α₁ subunits, α_1G (Ca_V3.1), α_1H (Ca_V3.2), and α_1I (Ca_V3.3), and their specific pharmacological properties have not yet been clearly identified (25). In this study, we systematically determined the effect of Zn²⁺ on transiently expressed HVA and LVA calcium channels and demonstrated that sensitivity to Zn²⁺ inhibition is α₁ subtype specific and may be dependent on selectivity filter residues outside of the EEEE/EEDD locus of the channels.

METHODS

Tissue culture and transient transfection

Human embryonic kidney tsA-201 cells were maintained in standard DMEM supplemented with 10% fetal bovine serum, 200 units/ml penicillin, and 0.2 mg/ml streptomycin at 37°C in a CO₂ incubator (26,27). Cells were split with trypsin-EDTA, plated on glass coverslips at 10% confluency, and allowed to recover for 12 h at 37°C. The cells were then transiently transfected with expression vectors containing cDNAs encoding wild-type or mutant calcium channel α₁, β_1b, and α₂-δ subunits and enhanced green fluorescent protein (eGFP) at a 1:1:1:0.2 molar ratio, using a standard Ca²⁺ phosphate protocol (26). All wild-type α₁ (α_1G, AF290212; α_1H, AF290213; α_1I, AF290214; α_1C, M67515; α_1A, M64373; α_1B, M92905; and α_1E, L15453), β_1b, and α₂-δ subunits cDNAs were generous gifts from Dr. Terry Snutch (University of British Columbia, Vancouver, British Columbia, Canada), and mutant α_1B subunit cDNAs were generous gifts from Dr. Gerald Zamponi (University of Calgary, Calgary, Alberta, Canada). After 12 h, the cell culture medium was replaced with fresh medium. The cells were allowed to recover for an additional 12 h and were subsequently kept at 28°C in 5% CO₂ for 1–2 days before physiological recordings were made.

Chemicals and solutions

All chemicals used in the cell culture were purchased from GIBCO (Invitrogen, Burlington, Canada). Chemicals used for physiological recordings were purchased from Sigma (St. Louis, MO).

Electrophysiology

Whole-cell patch-clamp (ruptured) recordings were performed using a MultiClamp 700A amplifier (Axon Instruments, Foster City, CA) linked to a personal computer equipped with pClamp9. Patch pipettes (Sutter borosilicate glass, BF 150-86-15) were pulled using a Sutter P-87 microelectrode puller and subsequently fire polished using a Narashige microforge. Pipettes (in the range of 2–4 MΩ) were filled with internal solution containing 108 mM Cs-methanesulfonate, 4 mM MgCl₂, 9 mM EGTA, and 9 mM HEPES (adjusted to pH 7.2 with TEA-OH). The cells were transferred to a 3.5-cm culture dish containing recording solution comprised of 20 mM BaCl₂ (or CaCl₂), 1 mM MgCl₂, 10 mM HEPES, 40 mM TEACl, 10 mM glucose, and 87.5 mM CsCl (adjusted to pH 7.2 with TEA-OH). Currents were elicited by stepping from a holding potential of −100 mV to various test potentials; Clampex software was used to control this process. Data were filtered at 1 kHz using a four-pole Bessel filter and digitized at a sampling frequency of 2 kHz.

Data analyses

All data were analyzed using Clampfit (Axon); curve fittings were carried out using SigmaPlot 4.0 (Jandel Scientific). Dose-response curves were fitted using the equation I/I₀ = 1/(1 + ([Zn²⁺]/IC₅₀)ⁿ), where I is the peak current response to a given test potential in the presence of Zn²⁺, I₀ is the current obtained in the drug-free condition (the control condition), [Zn²⁺] is the zinc concentration, IC₅₀ is the concentration at which 50% inhibition is obtained, and n is the Hill coefficient. Current-voltage relations obtained from peak current amplitude were fitted to the equation I = (G_max (V − V_r))/(1 + exp((V_h − V)/S)), where I is the measured peak current, G_max is the slope conductance, V is the test potential, V_r is the apparent reversal potential, V_h is the potential of half-maximal activation, and S is the slope of activation.

Statistics

Statistical analyses were carried out using SigmaStat 3.0 software (SPSS, Chicago, IL). The significance of differences between mean values from each experimental group was tested using a Student's t-test for two groups and one-way analysis of variance for multiple comparisons. Differences were considered significant if p < 0.05.

RESULTS AND DISCUSSION

Sensitivities of calcium channels to zinc inhibition are α₁ subunit specific

Differential inhibitory effects of Zn²⁺ on LVA calcium channels

We investigated the inhibitory effect of Zn²⁺ on three recombinant low-voltage–dependent T-type calcium channels (α_1G, α_1H, and α_1I) using a transient expression system. We evoked whole-cell currents in tsA-201 cells transiently expressing T-type calcium channels by a 10-ms depolarization step of −20 mV from a holding potential of −100 mV in 20 mM Ba²⁺. Representative current recordings for cells expressing Ca_V3.2 (α_1H) obtained for various Zn²⁺ concentrations are shown in Fig. 1 A1, and the corresponding time course of the development of the Zn²⁺ inhibition on the tail current is shown in Fig. 1 A2. Inhibition of the current was observed within 10 s of exposing cells to Zn²⁺ and rapidly reached equilibrium. Inhibition of the T-type calcium channel was reversible, as rapid recovery from Zn²⁺ inhibition was nearly complete after washout and removal of the Zn²⁺, consistent with previous reports for native neurons (14,19). Fig. 1 B shows a mean concentration-response curve fitted to the Hill equation (with a Hill coefficient of ∼1). As determined from eight independent experiments, the mean IC₅₀ value of Zn²⁺ for the Ca_V3.2 channel was 24.1 ± 1.9 μM.

FIGURE 1.

Differential block effect of Zn²⁺ on LVA calcium channels. (A) A typical representative of the dose-dependent blockade effect of Zn²⁺ on whole-cell current of Ca_V3.2 (α_1H) calcium channels. (A1) Current traces recorded in 20 mM Ba²⁺ for Ca_V3.2 channels during 10-ms voltage steps from a holding potential of −100 mV to −20 mV in the absence or presence of various Zn²⁺ concentrations, as indicated. Note that the current amplitude decreased with an increase in Zn²⁺ concentrations. (A2) The corresponding time course of development of block and recovery from Zn block in Ca_V3.2 channels is shown in A1. Comparison of the average dose-response curves (B) and IC₅₀ values (C) of Zn²⁺ block effect on three LVA calcium channels. The cells have been recorded from: Ca_V3.1 (α_1G), n = 7; Ca_V3.2 (α_1H), n = 8; and Ca_V3.3 (α_1I), n = 7. The data are presented as mean ± SE. (*) Statistical significance (p < 0.05) among the groups. The number in the parentheses in C indicates the number of cells used for the recordings.

The inhibitory effect of Zn²⁺ on Ca_V3.1 (α_1G) and Ca_V3.3 (α_1I) channels was determined using the same protocol as was used for the Ca_V3.2 channels (α_1H). Interestingly, these channels exhibited a different sensitivity to Zn²⁺, as the mean IC₅₀ values for Zn²⁺ varied depending on the α₁ subunit contained in the channel. Fig. 1 B shows a comparison of the mean concentration-response curves of Ca_V3.2 (α_1H) and Ca_V3.3 (α_1I) channels to the response curves for Ca_V3.1 (α_1G) channels. The mean concentration-response curve of the Ca_V3.2 channel shifted to a lower concentration than the response curves of the Ca_V3.1 or Ca_V3.3 channels (Fig. 1 B), indicating that the Ca_V3.2 channels are more sensitive to Zn²⁺ inhibition than are the other T-type calcium channels. The mean IC₅₀ values of Zn²⁺ for the three LVA channels are compared in Fig. 1 C. The IC₅₀ of Zn²⁺ for the Ca_V3.2 (α_1H) channels (24.1 ± 1.9 μM, n = 8) was significantly lower than that of the Ca_V3.1 (α_1G) (196.5 ± 50.4 μM, n = 7; p < 0.05) and Ca_V3.3 (α_1I) channel (152.2 ± 30.6 μM, n = 8; p < 0.05). The Hill coefficient for Ca_V3.2 (α_1H) was 1.3 ± 0.2, that for Ca_V3.2 (α_1G) was 1.1 ± 0.1, and the one for Ca_V3.3 (α_1I) was 1.0 ± 0.1. The differences of the Hill coefficients among three T-type channels were not statistically significant (p > 0.05), consistent with a recent report (28). These results demonstrate that the sensitivity to Zn²⁺ of the LVA calcium channels is intrinsically different among the subtypes, with Ca_V3.2-type channels being the most sensitive to the inhibitory effects of Zn²⁺. These findings are consistent with previous reports that α_1H has a higher sensitivity to Zn²⁺ than α_1G and α_1I (18,28). The Hill coefficients of Zn²⁺ for T-type channels varied ∼1, indicating a single binding site or no cooperative interaction between Zn²⁺ binding sites under the Ba²⁺ condition. The change in the potency of Zn²⁺ among the LVA channels likely results from the different binding affinities, assuming that Zn²⁺ binds to a similar site. Because the selectivity filter of all T-type channels has a conserved EEDD motif (29–31), the differing sensitivities, reflected by IC₅₀ values, of the T-type channels indicate that the Zn²⁺ inhibitory effect on these channels is likely regulated by the residues outside of the pore EEDD motif.

Differential inhibitory effects of Zn²⁺ on HVA calcium channels

We next performed whole-cell current recordings to determine the effect of Zn²⁺ on recombinant high- and intermediate-voltage–dependent calcium channels transiently expressed in tsA-201 cells. Similar to the results obtained for the LVA calcium channels, we found that Zn²⁺ inhibited the HVA calcium channels in a concentration-dependent manner (Fig. 2). Representative Ba²⁺ current recordings obtained for the L-type calcium channels (α_1C + β_1b + α₂ − δ) in various concentrations of Zn²⁺ are shown in Fig. 2 A, and the corresponding time course of the development for Zn²⁺ inhibition is shown in Fig. 2 B. We evoked currents by using a depolarization step of +10 mV from a holding potential of −100 mV in 20 mM Ba²⁺. Inhibition of the current developed within 10 s of exposure to Zn²⁺, rapidly reached equilibrium, and was eliminated promptly after washout of the Zn²⁺ (Fig. 2 A2), similar to the results obtained in experiments using transiently expressed T-type calcium channels (Fig. 1 B). Fig. 2 B shows the mean concentration-response curve fitted using the Hill equation (with a Hill coefficient of ∼1). As determined from eight independent experiments, the mean IC₅₀ value of Zn²⁺ inhibition of the L-type channel was 10.9 ± 3.4 μM.

FIGURE 2.

Differential block effect of Zn²⁺ on HVA calcium channels. (A) A typical representative of the dose-dependent blockade effect of Zn²⁺ on whole-cell current of Ca_V1.2 (α_1C) calcium channels. (A1) Current traces recorded in 20 mM Ba²⁺ for Ca_V1.2 channels during 10-ms voltage steps from a holding potential of −100 mV to −20 mV in the absence or presence of various Zn²⁺ concentrations, as indicated. Note that the current amplitude decreased with an increase in Zn²⁺ concentrations. (A2) The corresponding time course of development of block and recovery from Zn block in Ca_V1.2 channels is shown in A1. Comparison of the average dose-response curves (B) and IC₅₀ values (C) of Zn²⁺ block effect on four HVA calcium channels. The Hill coefficients to Zn²⁺ were Ca_V1.2 (α_1C), 0.88 ± 0.08 (n = 8); Ca_V2.1 (α_1A), 0.91 ± 0.06 (n = 8); Ca_V2.2 (α_1B), 0.92 ± 0.13 (n = 12); and Ca_V2.3 (α_1E), 0.55 ± 0.05 (n = 8). The data are presented as mean ± SE. Statistical significance (p < 0.05) (*) to Ca_V1.2 (α_1C) and (#) to Ca_V2.3 (α_1E) when one-way analysis of variance was used for comparison. The number in the parentheses in C indicates the number of cells used for the recordings.

The inhibitory effects of Zn²⁺ on N- (α_1B), P/Q- (α_1A), and R-type (α_1E) calcium channels (coexpressed with β_1b and α₂ − δ subunits) were determined using the same protocol used to determine the effect for L-type calcium channels (α_1C). Similar to T-type channels, HVA calcium channels responded differentially to the inhibitory effect of zinc, as the mean IC₅₀ values varied depending on the α₁ subunit. As depicted in the concentration-response curves in Fig. 2 B, the mean concentration-response curves for N- (α_1B), P/Q- (α_1A), and R-type (α_1E) calcium channels were shifted to higher concentrations compared to the response curve for the L-type calcium channel (Fig. 2 B), indicating that these channels are less sensitive to Zn²⁺ inhibition than is the L-type calcium channel. The IC₅₀ values of all four HVA calcium channels are compared in Fig. 2 C. The Zn²⁺ IC₅₀ value for the N-type (98.0 ± 17.9 μM, n = 12) and P/Q-type (110.0 ± 7.0 μM, n = 8) channels are ∼10-fold greater than that of the L-type (10.9 ± 3.4 μM, n = 8) channel. These differences are statistically significant (p < 0.05, Fig. 2 C). The Zn²⁺ IC₅₀ value for the R-type calcium channel was 31.8 ± 12.3 μM (n = 8), which is significantly higher than that of the L-type channel (p < 0.05) but lower than that of N- and P-/Q-type channels (p < 0.05). These results clearly demonstrate that the sensitivity to the Zn²⁺ inhibitory effect of the HVA calcium channels differs intrinsically with the highest sensitivity found in the L-type channels and the lowest sensitivity in the N- and P/Q- type channels. Again, as seen in LVA calcium channels, the Zn²⁺ effect on HVA calcium channels varied depending on the channel α₁ subtype, and this variation in sensitivity did not seem directly related to the selectivity filter EEEE residues, which are conserved in all HVA calcium channels (32,33).

Zn²⁺ block and the selectivity filter of calcium channels

The selectivity filter of the pore region of T-type channels has an EEDD locus (29–31), whereas the selectivity filter in high-voltage–gated channels has an EEEE locus (32,33). We found that the variation in sensitivity to Zn²⁺ block of the channels was independent of these conserved selectivity filter residues. Based on the observed Zn²⁺ IC₅₀ values for the various channels, we found that the order of sensitivity to the block by Zn²⁺ was approximately α_1C > α_1H > α_1E > α_1B = α_1A > α_1I = α_1G when Ba²⁺ was used as the charge carrier.

Most HVA calcium channels conduct Ba²⁺ ∼2-fold better than they conduct Ca²⁺ (see McDonald et al. (34)). This difference seems to result from a higher binding affinity for Ca²⁺ ions than for Ba²⁺ ions at the EEEE locus of the channel pore region (26,32,33,35). In contrast, the Ca²⁺ conductance of the LVA channels is similar to or slightly greater than that of Ba²⁺ (36–38). If Zn²⁺ inhibition of these channels depends on the charge carrier species and hence on the properties of the permeation pathway lining the pore region, the sensitivity of the HVA calcium channels to Zn²⁺ would be expected to be less in Ca²⁺ than in Ba²⁺, whereas that of T-type channels would remain unchanged. Therefore, we tested whether Zn²⁺ inhibition of these channels is affected by the charge carrier. Specifically, Ca_V2.2 N-type and Ca_V3.1 T-type calcium channels were compared because their IC₅₀ values in 20 mM Ba²⁺ were similar (100–200 μM), when the inhibitory effect of Zn²⁺ on Ba²⁺ and Ca²⁺ currents was measured. Fig. 3 A shows the I-V curves for both channels in Ba²⁺ (Fig. 3, A1 and A3) and Ca²⁺ (Fig. 3, A2 and A4) with or without 100 μM Zn²⁺. Consistent with previous reports (26,37–39), the current amplitude ratio between Ba²⁺ and Ca²⁺ (I_Ba/I_Ca) was ∼2 for the N-type channel and ∼0.8 for the Ca_V3.1 channel (Fig. 3, inset), and the I-V relations of the channels were not affected by Zn²⁺. As shown in Fig. 3 B, the IC₅₀ for Ca_V2.2 α_1B significantly increased (by ∼14-fold, p < 0.05) in 20 mM Ca²⁺ (1.21 ± 0.08 mM, n = 8) as compared to that in 20 mM Ba²⁺ (98.0 ± 17.9 μM, n = 12), suggesting that the current carrier species plays a role in regulating Zn²⁺ inhibition of the N-type channel. In contrast, the T-type Ca_V3.1 (α_1G) channel had a similar sensitivity to Zn²⁺ in Ba²⁺ (IC₅₀ 198.5 ± 50.4 μM, n = 7) and in Ca²⁺ (IC₅₀ 132.8 ± 21.2 μM, n = 7, consistent with Jeong et al. (18)); and the difference in the IC₅₀ values was not statistically significant (p > 0.05).

FIGURE 3.

Comparison of Zn effects on I_Ca and I_Ba of Ca_V2.2 (α_1B) N-type and Ca_V3.1 (α_1G) T-type calcium channels. (A) Representative current-voltage curves of Ca_V2.2 (α_1B) N-type and Ca_V3.1 (α_1G) T-type calcium channels in the absence or presence of 100 μM Zn²⁺: Ca_V2.2 (α_1B) channels (A1); 20 mM Ba²⁺ (A2); 20 mM Ca²⁺. Ca_V3.1 (α_1G) channels: (A3) 20 mM Ba²⁺; (A4) 20 mM Ca²⁺. (Inset) Comparison of the peak Ba²⁺ to Ca²⁺ current ratio of Ca_V2.2 (α_1B) and Ca_V3.1 (α_1G) calcium channels. (B) Summary of IC₅₀ values of Zn²⁺ on Ca_V2.2 (α_1B) and Ca_V3.1 (α_1G) channels. The blockade effects of Zn²⁺ on Ca_V2.2 (α_1B) channels are ∼10-fold higher in Ca²⁺ than in Ba²⁺ conditions, whereas those on Ca_V3.1 (α_1G) channels are similar. The data are presented as mean ± SE. from 12 cells expressing Ca_V2.2 (α_1B) channels and 6 cells expressing Ca_V3.1 (α_1G) channels. (*) Statistical significance (p < 0.05) between the different conditions. The number in parentheses indicates the number of cells used for the recordings.

Multiple mechanisms of Zn²⁺ block of VDCCs have been suggested, including pore block, surface charge screening, and gating modification (14,19,40); however, these mechanisms vary among the different channel subtypes. Our findings show that the charge carrier–dependent Zn²⁺ sensitivity cannot be explained by either hydration energy or surface charge screening alone. Experimental hydration energies for Ba²⁺, Ca²⁺, and Zn²⁺ are 298, 360, and 467 kcal/mol (41,42), respectively. If ion access to the pore required complete removal of the hydration shells, then these energies indicate that permeation/block follows the order Ba²⁺ > Ca²⁺ > Zn²⁺. Thus, Zn²⁺ block would be expected to be more sensitive to Ba²⁺ than to Ca²⁺. Although this seems to be the case for the Ca_V2.2 (α_1B) channel, it does not explain the differences seen in all the channels. For instance, all three T-type channels have similar Ba²⁺ to Ca²⁺ current ratios, ∼0.8, and the differences in Zn²⁺ sensitivities in the presence of Ca²⁺ and Ba²⁺ conditions are inconsistent. α_1I sensitivity was substantially higher in Ba²⁺ (IC₅₀ 152 μM; Fig. 1 D) than reported in Ca²⁺ (IC₅₀ 470 μM (18)); α_1H was lower in Ba²⁺ (IC₅₀ ∼ 24 μM; Fig. 1 D) than in Ca²⁺ (IC₅₀ 2.4 μM (18)); and in this study we showed that Zn²⁺ sensitivity of Ca_V3.1 (α_1G) was similar in Ba²⁺ and Ca²⁺ (Fig. 3). Surface charge screening also does not explain the differences in the Zn²⁺ effect between the subtypes. Mg²⁺ and Zn²⁺ are identically charged and contribute to surface charge screening equally. Mg²⁺ preferentially blocks Ba²⁺ currents more than Ca²⁺ currents through α_1G (43), whereas we showed that Zn²⁺ block of α_1G was comparable in Ba²⁺ and Ca²⁺. Thus, neither hydration energy nor surface charge screening alone appears to explain the influence of ion species–dependent modulation of channel conductance and Zn²⁺ block.

The selectivity filter of VDCCs alone does not explain the differences in the ion-dependent Zn²⁺ block effect. The molecular determinants of ion selectivity in VDCCs resides on the selectivity filter loci encoding EEEE/EEDD motif; however, α_1E encoded with EEEE locus seen in HVA channels, as an exception, exhibits a higher permeability to Ca²⁺ than to Ba²⁺, similar to LVA channels (38). Thus, additional residues other than the EEEE/EEDD motif are involved in ion selectivity of the pore. Previous studies in K⁺ (44) and Na⁺ (45) channels demonstrated that metal binding sites are dependent on the unique residues near the selectivity filter. Mutation of these residues substantially affected Zn²⁺ block. These findings bring into question whether Ca²⁺ channels may also contain specific extrapore sites regulating Zn²⁺ block.

Zn²⁺ blockade is regulated by the putative EF-hand extrapore region of N-type calcium channel

We previously reported that ion permeation in N-type channels is modulated by a putative EF-hand motif located in domain III near the EEEE locus (26). It contains a central glycine residue flanked by three acidic residues, reminiscent of the classical EF-hands of Ca²⁺ binding proteins (46,47). Disruption of this motif reduces the ability of the channel to distinguish between Ba²⁺ and Ca²⁺ ions without affecting the pore function. To determine whether this putative EF-hand motif is also involved in regulating Zn²⁺ inhibition of these channels, we compared the IC₅₀ values of Zn²⁺ obtained for wild-type N-type calcium channels to those obtained for two EF-hand mutant channels: a triple mutant in which all three negative charge residues in the putative EF-hand structure are replaced with positively charged residues (E1321K, D1323R, E1332R), and the G1326P mutant in which the central glycine is replaced with a proline. The latter mutant alters the ion permeability of the N-type channel by disrupting the EF-hand structure without changing the net local surface charge (26). We previously showed that these mutant channels exhibit biophysical properties similar to those of the wild-type channel (26). Representative recordings shown in Fig. 4 A reveal that Zn²⁺ inhibited the current activity of each mutant for both carriers, Ba²⁺ and Ca²⁺, in a dose-dependent manner. The mean Zn²⁺ IC₅₀ values were comparable between the wild-type (98.0 ± 17.9 μM, n = 12) and mutant channels in 20 mM Ba²⁺ (triple, 79.2 ± 13.9 μM, n = 8; G1326P, 60.0 ± 5.5 μM, n = 11); however, the mean Zn²⁺ IC₅₀ was significantly (p < 0.05) reduced in the mutant channels (wt, 1210 ± 80 μM, n = 8; triple, 290 ± 50 μM, n = 8; G1326P, 350 ± 70 μM, n = 11) when Ca²⁺ was used as the charge carrier (Fig. 4 B). Both the neutralization of three negatively charged residues and the replacement of central glycine with proline in the putative EF-hand region caused a similar effect on the Zn²⁺ sensitivity in Ca²⁺ condition, indicating that the effect is most likely independent of local surface charge screening. These results suggest that the putative EF-hand motif is involved in regulating Zn²⁺ block effect.

FIGURE 4.

Mutations of the putative EF-hand in domain III of the Ca_V2.2 (α_1B) N-type channel affected Zn²⁺ sensitivity. (A) Representative current traces from two mutants of Ca_V2.2 (α_1B) N-type calcium channels (trimutant, and G1326P) in the absence or presence of various concentrations of Zn²⁺, as indicated, in either 20 mM Ba²⁺ or 20 mM Ca²⁺. Trimutant: E1321K, D1323R, E1332R. (B) Summary of IC₅₀ values of Zn²⁺ on the wild-type and mutant Ca_V2.2 (α_1B) calcium channels. The data are presented as mean ± SE. (*) Statistical significance (p < 0.05) between the wild-type and mutant channels under the same recording conditions. The number in the parentheses indicates the number of cells used for the recordings.

The putative EF-hand in the domain III H5 loop of the N-type calcium channel α_1B subunit forms a helix-coil-helix structure (48). A molecular structure model suggests that it is capable of interacting with Ca²⁺, although its sequence exhibits some variation compared to the classical EF-hand. The putative EF-hand was previously shown to be involved in Ba²⁺ and Ca²⁺ permeability of N-type channels (26) as well as ω-conotoxin GVIA block (49). Eight of the HVA Ca²⁺ channel α₁ subunits appear to have this putative EF-hand motif (26), but none of the LVA Ca²⁺ channels, Na⁺ channels, and K⁺ channels have it in the homologous region (Fig. 5). Because mutation of the EF-hand motif of N-type channels increased Zn²⁺ sensitivity in Ca²⁺ toward that seen in α_1G T-type channels, it may be involved in distinguishing between Zn²⁺ block of the HVA and LVA channels. In the typical EF-hand motif, the −Z position glutamic acid residue contributes two carboxylate oxygen atoms to coordinate a Ca²⁺, and the central glycine is located at sharp bend position critical for Ca²⁺ coordination (46,47). Shown in Fig. 5, both −Z acidic residue (Ca_V2.1, E1376; Ca_V2.2, E1332; Ca_V2.3, E1289; Ca_V1.2, E1081) and the central glycine (Ca_V2.1, E1370; Ca_V2.2, E1326; Ca_V2.3, E1283; Ca_V1.2, E1072) are conserved among four HVA Ca²⁺ channels; however, the remaining residues vary among the channels. In comparison to the Ca_V2 α₁ subunits, Ca_V1.2/α_1C contains three additional residues, K1078, D1079, and G1080, between the central glycine (G1072) and the −Z glutamic acid (E1081), which may weaken the effectiveness of the putative EF-hand motif in coordinating Ca²⁺. This could explain our finding that Zn²⁺ sensitivity of Ca_V1.2/α_1C is higher than that of the Ca_V2 channels. Taken together, the presence and composition of the putative EF-hand motif may be partly responsible for the differences in Zn²⁺ block between HVA channels and LVA channels as well as between the HVA channels.

FIGURE 5.

Protein sequence alignments showing the extended pore regions of representative voltage-dependent calcium channels, sodium channels, and potassium channel. The box and asterisks indicate the identified selectivity filter residues. The solid line indicates the putative EF-hand motif region; solid circle, the conserved acidic residues and the central glycine; open circle, critical sites affecting Zn sensitivity of the channels. b, brain; s, skeletal muscle; and h, heart.

The differences in Zn²⁺ block effects on LVA channels do not appear to be explained by the putative EF-hand motif because it is absent in their α₁ subunit. In this study, we found that under Ba²⁺ conditions α_1H was ∼10-fold more sensitive to Zn²⁺ than α_1G and α_1I. In comparison, under Ca²⁺ conditions, α_1H was ∼100- to 200-fold more sensitive to Zn²⁺ than α_1G and α_1I (18,28). These charge carrier-dependent differences in Zn²⁺ sensitivities indicate that block is modulated by residues outside the EEDD selectivity filter locus. Similarly, α_1H has been shown to have higher sensitivities to Cu²⁺ and Ni²⁺ than the other T-type channels (18) through a mechanism that may involve noncharged residues. Cysteine or histidine pairs are known to be able to coordinate divalent ions and have been shown to affect ion selectivity when found in the pore regions of Na⁺ (45) and K⁺ channels (44). A recent report showed that the point mutation H191Q in the S3-S4 loop of domain I reduced Ni⁺ block of α_1H (50), indicating that H191 is critical for Ni⁺ block. Whether the high-sensitivity block of α_1H by Zn²⁺ also resides with this residue remains to be tested, but the study lends further support to the notion that extrapore residues can affect ion block of VDCCs.

In conclusion, we have systematically tested and compared the Zn²⁺ sensitivities of HVA and LVA calcium channels in a transient expression system. We show that Zn²⁺ block of VDCCs is dependent on the pore-forming α₁ subunit, but variation in the Zn²⁺ sensitivities of the channel subtypes is independent of EEEE/EEDD locus and is differentially regulated by permeant ions. Zn²⁺ block of the Ca_V2.2/α_1B N-type channel is modulated via a putative EF-hand motif outside of the EEEE locus. Although it is not clear how the EF-hand is involved in modulation of Zn²⁺ block, we envision a model in which Zn²⁺ binds at the pore of the channel, and other regions, such as the EF-hand, modulate the interaction between Zn²⁺ and the pore. Our study supports the notion that multiple mechanisms are involved in the channel subtype-dependent Zn²⁺ block. Further studies are required to identify other molecular determinants underlying the differences in the block effect of Zn²⁺ on the various calcium channel subtypes and to elucidate the mechanisms involving these determinants.