Complex modulation of Ca_v3.1 T-type calcium channel by nickel

Abstract

Nickel is considered to be a selective blocker of low-voltage-activated T-type calcium channel. Recently, the Ni²⁺-binding site with critical histidine-191 (H191) within the extracellular IS3–IS4 domain of the most Ni²⁺-sensitive Ca_v3.2 T-channel isoform has been identified. All calcium channels are postulated to also have intrapore-binding site limiting maximal current carried by permeating divalent cations (PDC) and determining the blockade by non-permeating ones. However, the contribution of the two sites to the overall Ni²⁺ effect and its dependence on PDC remain uncertain. Here we compared Ni²⁺ action on the wild-type “Ni²⁺-insensitive” Ca_v3.1_w/t channel and Ca_v3.1_Q172H mutant having glutamine (Q) equivalent to H191 of Ca_v3.2 replaced by histidine. Each channel was expressed in Xenopus oocytes, and Ni²⁺ blockade of Ca²⁺, Sr²⁺, or Ba²⁺ currents was assessed by electrophysiology. Inhibition of Ca_v3.1_w/t by Ni²⁺ conformed to two sites binding. Ni²⁺ binding with high-affinity site (IC₅₀ = 0.03–3 μM depending on PDC) produced maximal inhibition of 20–30 % and was voltage-dependent, consistent with its location within the channel’s pore. Most of the inhibition (70–80 %) was produced by Ni²⁺ binding with low-affinity site (IC₅₀ = 240–700 μM). Q172H-mutation mainly affected low-affinity binding (IC₅₀ = 120–160 μM). The IC₅₀ of Ni²⁺ binding with both sites in the Ca_v3.1_w/t and Ca_v3.1_Q172H was differentially modulated by PDC, suggesting a varying degree of competition of Ca²⁺, Sr²⁺, or Ba²⁺ with Ni²⁺. We conclude that differential Ni²⁺-sensitivity of T-channel subtypes is determined only by H-containing external binding sites, which, in the absence of Ni²⁺, may be occupied by PDC, influencing in turn the channel’s permeation.

Introduction

Low-voltage-activated (LVA or T-type) calcium channels constitute a separate group of plasma membrane voltage-gated calcium channels (VGCC) with biophysical properties enabling them to pass transient inward calcium current in response to low levels of depolarization (activation threshold around −60 mV) taking place from the relatively hyperpolarized cell’s resting membrane potential (V _r below −60 mV), which is required for the channels’ recuperation from steady-state inactivation occurring at higher V _r [1]. These channels are widely distributed in the central nervous system and in numerous peripheral tissues where they are believed to participate in the generation of neuronal low-threshold calcium spikes and cellular pacemaking activity [2], although a number of their other important physiological and pathological functions are demonstrated [3–6].

T-type calcium channels lack highly specific pharmacology [7, 8]. They are relatively insensitive to the classical organic (dihydropyridines, phenylalkylamines, benzothiazepines) and inorganic (Co²⁺, Cd²⁺) calcium antagonists, but could be effectively blocked by trivalent metal cations such as Y³⁺, La³⁺, and Gd³⁺ [9]. One of the most effective divalent cationic T-type calcium channel inhibitors was found to be Ni²⁺ [2]. However, pharmacology in general as well as Ni²⁺ sensitivity in particular of T-type VGCCs is a matter of significant regional, tissue-dependent variability. The cloning at the end of 90-s of three T-channel isoforms, Ca_v3.1, Ca_v3.2, and Ca_v3.3, encoded by different genes [10], and their tissue-specific expression provided molecular rational for such variability, as in functional tests only Ca_v3.2 demonstrated Ni²⁺ sensitivity in the micromolar range with IC₅₀ more than 20-fold lower compared to Ca_v3.1 and Ca_v3.3 [11].

Mutational analysis combined with functional tests of three T-channel isoforms led to the conclusion that high Ni²⁺-sensitivity is largely determined by the amino acid composition of S3–S4 extracellular loop of domain I (IS3–IS4) within the channel’s structure. It was found that in the event of Ca_v3.2, the presence of histidine-191 (H191) within this domain is critical for conferring high Ni²⁺-sensitivity [12], whereas relatively Ni²⁺-insensitive Ca_v3.1 and Ca_v3.3 in the equivalent positions contain glutamine. Mutation of the corresponding residue in Ca_v3.1 to histidine essentially enhanced Ni²⁺ inhibition, but still not to the level observed in the wild-type Ca_v3.2. Further studies have revealed that the binding site for nickel and other trace metals such as copper and zinc in Ca_v3.2 is composed of a Asp–Gly–His motif in IS3–IS4 and a second aspartate residue in IS2 [13]. Moreover, the ability of ascorbate to inhibit Ca_v3.2, but not Ca_v3.1 or Ca_v3.3 T-channels, was shown to result from metal-catalyzed oxidation of H191 [14].

However, the classical paradigm of VGCCs permeation and block postulates the presence within the channel’s pore of an additional binding site for divalent cations, which in the subclass of high-voltage-activated (HVA) VGCCs is formed by four glutamates (EEEE locus) and in T-type VGCCs by two glutamates and two aspartates (EEDD locus) belonging to the P-regions of respective channels’ pore-forming α1 subunits [15, 16]. This intrapore binding site is believed to limit the maximal current of permeating cations, Ca²⁺, Sr²⁺, and Ba²⁺, with the increase of their extracellular concentration, and to be responsible for channel’s blockade by non-permeating ones, which because of the strong binding and/or large size, stuck in the channel obstructing the passage of ions [17–20].

Implication of IS3–IS4 extracellular loop as critical determinant of T-channels blockade by Ni²⁺ reveals a second mode of Ni²⁺ action and largely ignores the possible contribution of intrapore binding site in the mechanism of the blockade. So far, the conclusion that Ni²⁺ may work on two sites, a relatively superficial low affinity one and the one located deeper within the pore, was reached based on the analysis of the kinetics and voltage dependence of Ni²⁺ action at high concentrations (i.e., above 0.1 mM) [21].

In the present work, we have compared by means of electrophysiology the concentration-dependence of blocking action of Ni²⁺ on the wild-type Ca_v3.1 (Ca_v3.1_w/t) and Ca_v3.1 with glutamine in the position of IS3–IS4 loop mutated to histidine (Ca_v3.1_Q172H) heterologously expressed in Xenopus oocytes. We find that concentration-dependence of Ni²⁺ block of both channels conforms to the existence of two binding sites for Ni²⁺, high- and low-affinity, and the affinity of both ones is being affected by mutation.

Materials and methods

Preparation of cRNA, isolation, maintenance, and injection of Xenopus oocytes

cDNAs of the rat wild-type Ca_v3.1, Ca_v3.1_w/t (GenBank/EBI accession number AF027984), and its Q172H mutant, Ca_v3.1_Q172H, were kindly provided by Dr. E. Perez-Reyes, University of Virginia. Both cDNAs were subcloned into pGEM-HEA plasmids, and capped Ca_v3.1_w/t and Ca_v3.1_Q172H complementary RNA (cRNA) was prepared in vitro from linearized plasmids using T7 mMessage mMachine in vitro transcription kit (Ambion).

Stage V and VI oocytes from adult female Xenopus laevis frogs were used for Ca_v3.1_w/t and Ca_v3.1_Q172H channels expression. The procedures for the oocytes isolation, maintenance, and injection did not differ from those detailed elsewhere [20]. The volume of the injected cRNA solution (0.2 μg/μl in water) was usually 50 nl per oocyte. The injection was performed using an automatic nanoliter-range injector (Drummond Scientific Company, USA). Oocytes were used for electrophysiological recordings on 4th–7th days after cRNA injection.

Electrophysiology and solutions

Membrane currents in the oocytes were recorded using a conventional double-microelectrode voltage-clamp technique using a TEV-200A amplifier (Dagan Corp., USA) as described before [20]. “Voltage recording” and “current passing” microelectrodes were pulled from borosilicate glass and had resistance of ~3 and ~1 MΩ, respectively, when filled with 3 M KCl. To reduce contamination of the expressed currents by the oocyte’s endogenous Ca²⁺-dependent currents, all measurements were conducted on oocytes injected with 50 nl BAPTA-KOH buffer (20 mM, pH = 7.4) 30 min prior to the electrophysiological experiment. The composition of the extracellular solution for the recording of the expressed T-type channel currents was the following (in mM): 10 CaCl₂ (or BaCl₂ or SrCl₂), 107.5 TEA-Cl, 5 HEPES, pH 7.4 (adjusted with TEA-OH). Ni²⁺ was added to this solution to the required concentration as NiCl₂ × 6H₂O. Rapid (within a few seconds) exchange of the solutions around the oocyte under electrophysiological recording was achieved by its sequential immersion into the chambers filled with different solutions, as described in http://www.biph.kiev.ua/departments/shuba/research_files/solutions.htm. All reagents used in the study were from Sigma.

Data analysis and statistics

Currents at each condition were measured in several oocytes, normalized to the maximal control current in the same oocyte and then averaged. The data points were presented as mean ± SEM (standard error of the mean). Paired Student’s t test was used to determine statistical significance of the differences with p < 0.05 was considered as statistically significant. Concentration–response relationships were constructed by plotting the fraction of Ni²⁺-inhibited current (FIC), FIC = (I_Ctrl−I_test)/I_Ctrl, where I_Ctrl is the current in the absence of Ni²⁺ and I_test is the current in the presence of given Ni²⁺ concentration, against Ni²⁺ concentration. The experimental data points in concentration–response dependencies represent mean ± SEM FIC values for each Ni²⁺ concentration derived from different oocytes with “n” representing the number of oocytes tested. To obtain as precise as possible concentration–response relationships, we used a wide range of Ni²⁺ concentrations (from 1 nM to 3 mM) enabling correct determination of the onset and saturation of the block, and “n” per data point not less than 4 when the data point’s position caused little or no doubt and up to 12 for the data point with slightest uncertainty. Numerical values for the parameters of concentration–response dependencies were obtained from the fits of the data points with Hill function. Standard errors of the parameters were provided by fitting software, and are indicative of how well the curve fits the data. Data analysis was performed using pCLAMP 8 (Axon Instr., Foster City, CA, USA) and Origin 7.0 (Microcal, Northampton, MA, USA) software.

Results

Concentration–response relationship for Ni²⁺ action

Divalent cations compete with each other for binding, thus, one of the objectives of our study was to examine how different charge carries, Ca²⁺, Sr²⁺, or Ba²⁺, influence the action of Ni²⁺. Figure 1 shows original recordings of calcium (I _Ca), strontium (I _Sr) and barium (I _Ba) currents through Ca_v3.1_w/t and Ca_v3.1_Q172H channels at 10 mM of the respective charge carrier. Application of 10 μM Ni²⁺ inhibited all currents. Surprisingly, the extent of inhibition for both Ca_v3.1_w/t and Ca_v3.1_Q172H channels as well as for different charge carriers appeared quite similar, ranging from 20 to 30 %, despite the two channels differed in critical histidine within the IS3–IS4 loop.

Fig. 1.

Concentration-dependence of nickel action. a, b, c Concentration-dependencies for the blocking action of Ni²⁺ on Ca_v3.1_w/t channel; data points—experimental values (mean ± SEM, n = 4–12) for the fraction of inhibition of Ca²⁺ (a, filled circles), Sr²⁺ (b, filled squares) and Ba²⁺ (c, filled triangles) currents elicited by step depolarization to V _m = −30 mV; corresponding open symbols represent data points for the low-affinity block only; smooth lines—fit of the experimental data points with the sum of two Hill functions (solid line), and the dissection of the overall effect on the constituent Hill functions describing high- (dotted line) and low-affinity (dashed line) blockade; the parameters of the each Hill function: the concentration of half maximal inhibition, IC₅₀, maximal extent of inhibition, A _max, and cooperativity coefficient, p, are presented in the plots; insets in each panel show representative recordings of the corresponding current under steady-state inhibition by specified Ni²⁺ concentrations (shown on left). d, e, f Same as in a, b, c, respectively, but for Ca_v3.1_Q172H channel

To exclude the possibility that such similarity is just a coincidence for one Ni²⁺ concentration, we have acquired the concentration–response relationships for the blocking action of Ni²⁺ on I _Ca, I _Sr, and I _Ba through Ca_v3.1_w/t and Ca_v3.1_Q172H. Since the maximum of the inward currents through both channels irrespective of the charge carrier was observed at membrane potential (V _m) around −30 mV, we selected this potential for measuring concentration–response relationship.

Consecutive applications of ever-increasing concentrations of Ni²⁺ caused progressive inhibition of all currents (Fig. 1). However, the extent of inhibition increased non-monotonically with concentration. We observed distinct current inhibition of about 10 % through both channels at extracellular Ni²⁺ concentration ([Ni²⁺]_o) as low as 0.1 μM. At higher [Ni²⁺]_o in the range of 1-10 μM the extent of inhibition saturated at about 20–30 % and than increased again to reach 100 % at concentrations of Ni²⁺ above 1 mM (Fig. 1). Such non-monotonic concentration-dependence of Ni²⁺ action was characteristic of both Ca_v3.1_w/t and Ca_v3.1_Q172H channels and all charge carriers, indicating the presence of at least two binding sites with different affinity for Ni²⁺ through which it induces blocking effects.

Despite similarity in overall appearance of concentration dependencies, the experimental data points for each channel and charge carrier fell in the different concentration ranges, suggesting that the presence of critical histidine in the IS3–IS4 loop as well as the nature of permeating cation influence Ni²⁺ binding affinity. To determine quantitative parameters of Ni²⁺ blockade, i.e., concentration of half maximal inhibition, IC₅₀, maximal extent of inhibition, A _max, and cooperativity coefficient, p, we have fitted the experimental data points with Hill functions. Under no circumstances was a single Hill function sufficient to adequately describe experimental data: the best fit of each concentration dependency required two Hill functions (Fig. 1) of which the first one described the channels’ blockade via Ni²⁺ interaction with high-affinity binding site and the second one—with low-affinity binding site (see Table 1). Concentration of half maximal inhibition, IC₅₀, associated with high-affinity binding site was in the range of 0.031 μM (I _Ba, Ca_v3.1_Q172H) to 3.2 μM (I _Ca, Ca_v3.1_w/t) with maximal extent, A _max, varying between 22 % and 27 % (Table 1). IC₅₀ for the low-affinity inhibition had more than 100-fold higher value and strongly depended on the presence of critical histidine in the IS3-IS4 loop decreasing in charge carrier-specific manner from 242 to 702 μM for Ca_v3.1_w/t to 120–161 μM for Ca_v3.1_Q172H. Low-affinity inhibition had higher efficacy compared to the high-affinity one being responsible for the 73–80 % of the overall Ni²⁺ effect. In all cases, the Hill coefficient, p, was close to “1”, suggesting non-cooperative Ni²⁺ binding to both sites.

Table 1.

Parameters of high-affinity and low-affinity blockade of Ca²⁺, Sr²⁺, and Ba²⁺ currents through Ca_v3.1_w/t and Ca_v3.1_Q172H channels by Ni²⁺, as determined from the fit of experimental data points for concentration–response dependencies with Hill functions

Ion	Cav3.1w/t						Cav3.1Q172H
	IC50 (μM)		A max (%)		p		IC50 (μM)		A max (%)		p
	High	Low	High	Low	High	Low	High	Low	High	Low	High	Low
Ca2+	3.2 ± 0.61	611 ± 75	27 ± 1.8	73 ± 2.9	1.00 ± 0.20	0.83 ± 0.08	0.11 ± 0.05	120 ± 12	21 ± 1.5	79 ± 4.1	0.89 ± 0.23	0.89 ± 0.10
Sr2+	0.091 ± 0.01	242 ± 17	26 ± 2.0	74 ± 1.8	1.11 ± 0.11	0.83 ± 0.04	0.063 ± 0.01	161 ± 15	23 ± 1.0	77 ± 2.4	1.07 ± 0.26	1.00 ± 0.09
Ba2+	0.031 ± 0.008	702 ± 55	22 ± 1.9	78 ± 2.1	1.00 ± 0.23	0.91 ± 0.06	0.013 ± 0.006	130 ± 14	20 ± 1.1	80 ± 2.8	1.00 ± 0.31	1.13 ± 0.13

Standard errors of the parameters were provided by fitting software, and are indicative of how well theoretical Hill curve fits experimental data

IC ₅₀ concentration of half maximal inhibition, A _max maximal extent of inhibition, p cooperativity (Hill) coefficient

IC₅₀ for the high-affinity Ni²⁺ block of both Ca_v3.1_w/t and Ca_v3.1_Q172H to a greater extent depended on the permeating ion species than the one for the low-affinity block: it was the smallest when Ba²⁺ was used as a charge carrier consecutively increasing for Sr²⁺ and Ca²⁺ (Table 1). This suggests essentially stronger competition of Ca²⁺ compared to Sr²⁺ and even more so to Ba²⁺ with Ni²⁺ for high-affinity site and is generally consistent with the Ca²⁺>Sr²⁺>Ba²⁺ order of apparent binding affinities inside calcium channels during permeation [17].

Voltage dependence of Ni²⁺ action

The parameters of Ca_v3.1_w/t and Ca_v3.1_Q172H channels blockade suggest that Q172H mutation affects the affinity of Ni²⁺ binding at both high- and low- affinity sites. In accordance with previous studies [11, 12], we hypothesize that the high-affinity site is formed by the EEDD locus within the permeation path, whereas the low-affinity site is situated in the extracellular IS3-IS4 loop and must include critical histidine for enhanced affinity. Our data indicate that the binding of Ni²⁺ to the extracellular low-affinity site can, on the one hand, stabilize the channel’s closed states via impairment of conformational changes associated with gating, as was proposed in the original study [12], and, on the other hand, allosterically modulate Ni²⁺ interaction with the intrapore EEDD locus. EEDD locus is located inside the channel’s pore and, therefore, interaction of Ni²⁺ with it is believed to be voltage-dependent [22]. At the same time, voltage dependence of Ni²⁺ action at low-affinity site is hard to predict, as binding per se may be voltage-independent (because the site is located at the extracellular surface), but conformational changes associated with it may depend on membrane voltage. Besides, at concentrations of Ni²⁺ that engage low-affinity site in the blocking action, the channel is already partially inhibited via the high-affinity site, which makes appreciation of the true voltage dependence of the blockade at low-affinity site impossible.

Thus, in our experiments, we could only evaluate the voltage dependence of high-affinity blockade. To do so, we selected 10 μM Ni²⁺ concentration, which irrespective of the charge carrier, on the one hand, was able to produce maximal inhibition of both Ca_v3.1_w/t and Ca_v3.1_Q172H channels at high-affinity site, but, on the other hand, was yet insufficient to produce appreciable block at low-affinity site.

Figure 2 shows that for all charge carriers the action of 10 μM Ni²⁺ on both channels was voltage-dependent. In all cases, at V _m just above activation threshold, Ni²⁺ even potentiated the current, which turned to current inhibition at more depolarized V _m. At depolarization up to V _m = −0 ± 10 mV, the fraction of inhibited current remained nearly constant, constituting around 20 % with the tendency to decrease at higher voltages. Unfortunately, with the decrease of inward current through the channels, as depolarizations approached the apparent reversal potential (V _rev), a huge uncertainty in determining the fraction of inhibited current occurred. This situation was complicated by the slight shift of the V _rev in the presence of Ni²⁺, which altogether made correct assessment of voltage dependence at potentials above +20 mV highly unreliable. Nevertheless, the general conclusion, which can be drawn from these studies is that at 10 μM Ni²⁺ even potentiates the current through the channels at very negative V _m below −60 to −40 (depending on the channel and charge carrier), produces maximal inhibition at V _m between −30 mV and 0 mV, after which the extent of inhibition decreases again. Such voltage dependence is characteristic for the blockade of sodium currents through HVA and LVA calcium channels by Ca²⁺ [23, 24] and is generally consistent with the location of the binding site for the blocking ion sufficiently inside the channel where it experiences the change in the transmembrane voltage.

Fig. 2.

Voltage dependence of nickel action. a, b, c Lower part of each plot—normalized and averaged I–V relationships (mean ± SEM, n = 6–8) of Ca²⁺ (a, circles), Sr²⁺ (b, squares), and Ba²⁺ (c, triangles) currents through Ca_v3.1_w/t channels under control conditions (Ctrl, filled symbols) and in the presence of 10 μM Ni²⁺ (open symbols); solid lines represent the best fit of experimental data points with the product of Boltzmann and Goldman-Hodgkin-Katz equations; the parameters of the Boltzmann fits: the half-maximal activation potential, V _1/2, and the slope factor, k, are shown in the plots; * statistically significant differences (p < 0.05) in the currents’ amplitude. Upper part of each plot—the fraction of Ni²⁺-inhibited current derived from experimental data points of the I–Vs (gray symbols) and from the fits of the I–Vs (solid lines). d, e, f Same as in a, b, c, respectively, but for Ca_v3.1_Q172H channel

Notably, the fit of experimental data points of the I-Vs with the product of Boltzmann and Goldman-Hodgkin-Katz (GHK) equations (Fig. 2) has shown that in the presence of 10 μM Ni²⁺ the voltage of half-maximal activation, V _1/2, of either Ca_v3.1_w/t or Ca_v3.1_Q172H channel experienced 1–2 mV hyperpolarizing shift. Although this change did not reach the level of statistical significance, it indicated that binding of the Ni²⁺ to high-affinity site may slightly modify the channel’s activation gating. At the same time, hyperpolarizing shift of V _1/2 for Ca_v3.1_Q172H compared to Ca_v3.1_w/t, which reached maximal value of 7 mV in the presence of Sr²⁺ as charge carrier, suggested that mutation per se is allosterically coupled to channel activation gating.

Discussion

In this study, we for the first time provide direct experimental evidence on the existence of two binding sites through which nickel interacts with T-type calcium channels to induce their inhibition. Before, the existence of such sites for Ca_v3.1 was predicted based on the analysis of kinetics and voltage dependence of Ni²⁺ action [21], but not actually demonstrated from concentration dependencies. For the commonly considered to be Ni²⁺-insensitive Ca_v3.1_w/t channel, we find that the two sites differed by 2-to-4-orders in Ni²⁺ affinity (depending on charge carrier), which allowed to classify them as high and low affinity, and by 2.5-to-3.5-fold in the channel blocking efficacy.

The high-affinity site is obviously located inside the Ca_v3.1_w/t channel and is most likely formed by the EEDD locus of the selectivity filter. The apparent affinity of Ni²⁺ binding with this site is characterized by the value of IC₅₀ in the submicromolar range, and is dependent on the permeating ion species, indicating competition between blocker and charge carrier for the binding. Despite potent binding, the efficacy of the channel’s blockade by Ni²⁺ via interaction with EEDD locus is quite low, not exceeding 30 %. Thus, the binding of Ni²⁺ within the T-type calcium channel does not essentially impair the passage of permeating cations. This contrasts the situation taking place in the HVA subfamily of VGCCs in which cationic blockers like Cd²⁺ via interaction with the channel’s EEEE selectivity locus block the channel completely. The first-order explanation for this difference may come from the fact that the side chain in aspartate (i.e., D) is one methylene group shorter than in the glutamate (i.e., E) making the pore of the T-type calcium channel wider. However, probing the pore with large organic cations has shown that the pore size in Ca_v3 channels is in fact the tiniest amongst the all types of VGCCs [25]. Previous studies of VGCCs’ block by polyvalent cations have established that (1) the blocker can equally access the blocking site in the closed and open channel [26, 27] and that (2) the interaction with the same single site within permeation pathway is responsible for both tonic and use-dependent blocking effects [27]. However, separation of the concentration–response dependencies onto multiple components enabling to reveal only partial high-affinity blockade of VGCCs so far has not been performed. Since such blockade cannot be explained in the framework of the simple interaction model in which the blocking ion binds to a single high-affinity site occluding the pore for the single-file passage of permeating cations, one has to postulate more complex interactions involving several intrapore sites with different affinity, as was proposed previously [27, 28]. Although additional studies are needed to come up with the most adequate model, tentatively one can suggest that the high-affinity site within the Ca_v3.1 permeation pathway is in fact not a single one, but a composite one consisting of the two adjacent parts separated by an insignificant barrier (Fig. 3). Its outer, low-affinity part is located at the extracellular side of the pore and bears “permissive” function. The latter is followed by the deeper, high-affinity part, which when occupied by Ni²⁺, blocks the channel. Both parts of the composite site are accessible in closed and open channel. In the assumption of a single-file stair-step mechanism, Ni²⁺ can occupy the deeper, “blocking” site and block the channel only by jumping to it from the outer “permissive” one, but because of the low affinity of the latter the blocker can be easily competed out from the outer site by permeating cations, which would thus limit the access of Ni²⁺ to the deeper site and reduce the blocking effect.

Fig. 3.

Tentative model of Ni²⁺ action on Ca_v3.1 channel. The extracellular low-affinity Ni²⁺-binding site modulates channel permeation, inhibition, and gating allosterically, and its affinity for Ni²⁺ depends on the presence of critical histidine. The high-affinity, but low-efficacy component of channel blockade by Ni²⁺ may be likely determined by the presence in the channel’s permeation pathway of a composite Ni²⁺ binding site consisting of the two adjacent parts separated by an insignificant barrier, the low-affinity one located at the extracellular side of the pore, which bears “permissive” function, and the deeper, high-affinity one which when occupied by Ni²⁺ blocks the channel. Both parts of the composite site are accessible in the channel’s open and closed states. When Ni²⁺ occupies the deeper, “blocking” site, it blocks the channel (left panel). However, Ni²⁺ can reach the “blocking” site only by jumping to it from the outer “permissive” one, but because of the low affinity of the latter the blocker can be easily competed out from the outer “permissive” site by permeating cations (i.e., Ca²⁺, Sr²⁺ or Ba²⁺, right panel), which would thus limit the access of Ni²⁺ to the deeper site and reduce the blocking effect. The affinity of the ion to the site is schematically depicted by the number of small circles coming into the contact with the ion. Single-headed thick arrows indicate the direction of ion transfer, double-headed arrows—competition between ions

The block of Ca_v3.1_w/t channel by 10 μM Ni²⁺ appeared voltage-dependent decreasing at V _m below −40 mV and above +20 mV. Moreover, progressive relieve of the block at negative V _m at some point even turned to current potentiation. One explanation to this may be given if to suggest that at these potentials the driving force for Ni²⁺ may be high enough to relieve it from binding and to cause its penetration into the cell making it a current carrier. However, given only 10 μM Ni²⁺ concentration, in order to generate comparable to Ca²⁺, Sr²⁺, or Ba²⁺ current, Ni²⁺ would have to have 100 to 1,000-times-higher transfer rate than permeating cation, which is hardly possible. Most likely, the apparent current potentiation at negative potentials results from some effect of the high-affinity Ni²⁺ binding on the channel’s activation gating. Indeed, we have detected 1–2 mV hyperpolarizing shift of V _1/2 for channel activation in the presence of 10 μM Ni²⁺, which although statistically insignificant, nevertheless could account for the observed current potentiation. Even though the Ni²⁺ action at voltages close to V _rev and higher appeared hard to assess experimentally, consistent with previous studies [23, 24, 26, 27], the channel block most likely becomes relieved, as these voltages do not promote Ni²⁺ entering the channel and reaching its high-affinity binding site.

The IC₅₀ of high-affinity Ca_v3.1_w/t channel inhibition by Ni²⁺ was the smallest when Ba²⁺ was carrying the current consecutively increasing for Sr²⁺ and Ca²⁺. This is consistent with the classical paradigm of VGCCs permeation and block when blocking and charge-carrying ions compete for the same intrapore binding site with the least competition produced by Ba²⁺, and the most by Ca²⁺ with Sr²⁺ being in between.

Q172H mutation only slightly affected the parameters of high-affinity Ca_v3.1_Q172H channel blockade by Ni²⁺ compared to the Ca_v3.1_w/t one with essential change in IC₅₀ observed only for the Ca²⁺ current (see Table 1). This suggests that low-affinity Ni²⁺ binding to the extracellular site within S3–S4 loop of domain I of which histidine (i.e., H) is a critical component and/or conformation of this site depending on whether or not histidine is present exerts only weak allosteric influence on permeation pathway, where high-affinity Ni²⁺ binding site is located. This is consistent with the original hypothesis that the binding of Ni²⁺ to the extracellular low-affinity site inhibits channel via stabilization of the closed states due to impairment of channel gating [12].

Q172H mutation increased the affinity of Ca_v3.1_Q172H channel blockade by Ni²⁺ at extracellular low-affinity binding site by up to fivefold (depending on charge carrier, Table 1). Despite such an increase, the respective IC₅₀ still remained in the range of lower hundreds of μM, which is about four orders bigger than IC₅₀ of high-affinity blockade. At the same time, the efficacy of low-affinity blockade, which for Ca_v3.1_w/t reached as much as 73–78 % (depending on charge carrier), exceeding that of high-affinity one by 2.7–3.5-fold, was hardly changed by the mutation (i.e.,77–80 %, Table 1). Much smaller efficacy of high-affinity blockade of Ca_v3.1 channel combined with quite low Ni²⁺ concentrations at which it becomes evident (i.e., below 1 μM) probably explains why it was overlooked in previous studies. The fact that despite the presence of histidine, the IC₅₀ of low-affinity Ni²⁺ blockade of Ca_v3.1_Q172H channel still remained about 12-fold higher than for Ca_v3.2 channel (see [12]) suggests that either there are additional factor(s) determining Ni²⁺ block or the efficacy Ca_v3.2 channel blockade at high-affinity site is much stronger compared to Ca_v3.1 one, such that it makes a major contribution to the overall Ni²⁺ effect. The second possibility seems to be most likely, as in the only study in which biphasic Ni²⁺ action on Ca_v3.2 channel (expressed in HEK-293 cells with Ca²⁺ as charge carrier) was reported, the fraction of high-affinity blockade (IC₅₀ ~ 1.9 μM) reached 60–70 % of the overall effect with the low-affinity one (IC₅₀ ~ 1.35 mM) being responsible for only 30–40 % [29].

The affinity of Ni²⁺ binding at extracellular low-affinity site of both Ca_v3.1_w/t and Ca_v3.1_Q172H channels also depended on permeating ion species, indicating that Ca²⁺, Sr²⁺ and Ba²⁺ differentially compete with Ni²⁺ for the binding irrespective of whether histidine is part of it. Nevertheless, the presence of histidine not only increased the binding affinity of Ni²⁺, but also changed the order of competition for permeating ions.

The presence of the external binding site for divalent cations, which was initially postulated to participate in the control of Na⁺–Ca²⁺ permeability switch of HVA calcium channels in mollusc neurons [30], seems to be a common feature of all types of calcium channels, although its locations and functional significance appears to be more complex and diverse. Except for the IS3–IS4 loop of LVA Ca_v3.2 channels (T-type) in which it largely determines Ni²⁺ sensitivity [12, 13], it was identified in the extracellular IIIS5-P region of HVA Ca_v2.2 (N-type) channel where it was implicated in the control of differential Ca²⁺ and Ba²⁺ permeability [31]. Yet another extracellular domain, IS5–IS6 linker, was shown to participate in determining single-channel conductance of the subtypes of HVA L-type Ca_v1.1 and Ca_v1.2 channels [32]. Finally, just recently it was demonstrated that the trace metals zinc and copper at physiologically relevant concentrations reduce voltage sensitivity of HVA Ca_v2.3 channel (R-type) activation by occupying the site located at the interface of the channel’s IS1-IS2 and IS2-IS3 loops, the effect which may underlie the modulation of synaptic transmission and plasticity in the brain by trace metals [33].