Gadolinium inhibits Na⁺-Ca²⁺ exchanger current in guinea-pig isolated ventricular myocytes

Abstract

The trivalent cation, gadolinium (Gd³⁺) is commonly used to inhibit stretch-activated channels. In this report, we show that Gd³⁺ also inhibits ionic current (I_NaCa), carried by the Na⁺-Ca²⁺ exchanger protein. Under selective recording conditions, Gd³⁺ inhibited both outward and inward I_NaCa from guinea-pig isolated ventricular myocytes in a dose-dependent manner, with half-maximal inhibition concentrations (IC₅₀) of 30.0±4.0 μM at +60 mV (Hill-coefficient, h=1.04±0.13) and 20.0±2.7 μM at −100 mV (h=1.13±0.16), respectively (P>0.05, n=5–9). Thus, inhibition was not voltage-dependent. The time from Gd³⁺ application to steady-state effect was slow compared to the divalent blocker Ni²⁺. The slow time course appeared to reflect gradual Gd³⁺ accumulation at its binding site on the exchanger, rather than a use-dependent blocking mechanism. This study indicates that for experiments in which Gd³⁺ is used, its inhibitory effect on I_NaCa should be taken into account.

Introduction

Gadolinium (Gd³⁺) is a trivalent lanthanide that has been widely used as a stretch-activated channel (SAC) blocker in a wide range of tissue types (Yang & Sachs, 1989). Some reports, however, indicate that the actions of Gd³⁺ are not restricted to SACs. In ventricular cardiomyocytes, Gd³⁺ blocks L-type Ca current (I_Ca,L; Lacampagne et al., 1994) and the rapid delayed rectifier K current (I_Kr; Pascarel et al., 1998), whilst not inhibiting stretch-induced increase of intracellular Ca²⁺ (Hongo et al., 1997) or SAC in atrial cells (Zhang et al., 2000). Hongo et al. (1997) suggested that the lack of effect of Gd³⁺ on stretch-induced increases in [Ca²⁺]_i may result from an additional action: Gd³⁺ might inhibit Ca²⁺ extrusion on the sarcolemmal Na⁺-Ca²⁺ exchanger, as this plays a vital role in Ca²⁺ homeostasis (Blaustein & Lederer, 1999).

The exchanger is expressed widely in mammalian tissues e.g. heart, brain, kidney, colon, spleen, lung and pancreas (Blaustein & Lederer, 1999). Thus if Gd³⁺ can block the exchanger this would be of relevance to studies on a variety of experimental systems. The stoichiometry of the Na⁺-Ca²⁺ exchange process (3Na⁺ ions transported per 1 Ca²⁺ ion) means that it generates an ionic current (I_NaCa; Kimura et al., 1986; 1987). The aim of the present study was to utilize direct measurements of I_NaCa from guinea-pig ventricular myocytes to determine whether or not Gd³⁺ does inhibit the exchanger.

Methods

Cell isolation

Single ventricular myocytes from hearts of male guinea-pigs (400–600 g) were isolated as previously described (Hobai et al., 1997). The cells were kept at 4°C in high-K⁺, low Cl⁻ storage medium (Kraft-Brühe, KB medium) containing (in mM): L-glutamate 100, KCl 30, Na-pyruvate 5, taurine 20, creatine 5, succinic acid 5, Na₂ATP 2, β-OH butyrate 5, glucose 20, MgCl₂ 5, EGTA 1, HEPES 10 (pH adjusted to 7.2 with KOH).

Voltage-clamp technique

An aliquot of cell suspension was allowed to settle on the glass bottom of a Perspex chamber on the stage of an inverted microscope (Nikon Diaphot) for several minutes. Then, the bath was continuously superfused at 37°C with Tyrode's solution (see below). Whole-cell patch clamp experiments were performed using an Axopatch 200A amplifier (Axon instruments, U.S.A.). Patch pipettes (Corning 7052 glass, AM Systems) were pulled using a Flaming/Brown P87 puller and fire-polished to final resistance of 1–2 MΩ (Narishige MF 83 microforge, Japan). Protocols were generated and data recorded on-line with p-Clamp 6.0 software via an analogue-to-digital converter (Digidata 1200B, Axon instruments, U.S.A.). Data are expressed as mean±s.e.mean; t-test and ANOVA were used for statistical analysis. P values less than 0.05 were taken as significant.

Solutions

Tyrode's solution contained (in mM): NaCl 140, HEPES 5, glucose 10, KCl 4, CaCl₂ 2.5, MgCl₂ 1 (pH adjusted to 7.45 with NaOH). We used similar solutions to Hinde et al. (1999) to measure I_NaCa. The extracellular solution for recording I_NaCa was K⁺-free Tyrode's solution containing 10 μM strophanthidin, 10 μM nifedipine and 1 mM BaCl₂ to eliminate K, Ca, background and Na-K pump currents. Solutions were applied using a home-built rapid solution application device. Intracellular solution contained (in mM): Cs-aspartate 113, NaCl 20, MgCl₂ 0.4, Tris-ATP 5, HEPES 10, glucose 5, BAPTA 10, tetraethylammonium (TEA) 20, CaCl₂ 1, pH 7.2 (titrated with CsOH). The combination of 10 BAPTA and 1 CaCl₂ gave a free pipette Ca concentration of 20 nM (calculated with the Maxchelator program).

Chemicals

Gadolinium chloride was purchased from Sigma and dissolved in the external I_NaCa solution to the concentrations shown in ‘Results'.

Results

The solutions described above have previously been demonstrated to allow I_NaCa to be measured as current sensitive to external Ni²⁺ or 0Na/0Ca solution (Hinde et al., 1999). To determine the effects of Gd³⁺, we applied descending ramp pulses every 12 s from +80 mV to −120 mV (dV/dt=0.4 V s⁻¹) from a holding potential of −80 mV. Representative net currents before and after Gd³⁺ are shown in Figure 1A inset. Both outward and inward currents were inhibited by 100 μM Gd³⁺. The Gd³⁺-sensitive current showed an outwardly rectifying current-voltage (I–V) relationship with a reversal potential (E_rev) close to −60 mV (Figure 1A). The observed E_rev was higher than that calculated for I_NaCa using the free [Ca] and [Na] values for the pipette solution given in the Methods (which was in excess of −150 mV, using equation 6′ of Blaustein & Lederer, 1999). It was, however, consistent with recent data (Convery & Hancox, 1999), which suggest that buffering of subsarcolemmal and bulk [Ca] can differ under selective recording conditions for I_NaCa. The characteristics of the Gd³⁺-sensitive current, therefore, resembled those reported previously for I_NaCa (e.g. Convery & Hancox, 1999; Hinde et al., 1999). To confirm the identity of the Gd³⁺-sensitive current, extracellular Na⁺ and Ca²⁺ were replaced with Li⁺ and Ba²⁺ (0Na/0Ca solution to abolish I_NaCa). With the exchanger inhibited, the residual current was not significantly affected by Gd³⁺ (Figure 1B). Similar results were obtained when I_NaCa was completely blocked with the widely used blocker of I_NaCa Ni²⁺ (10 mM; data not shown). Collectively, these data showed that the current component inhibited by Gd³⁺ was I_NaCa.

Figure 1.

The effect of 100 μM Gd³⁺ on I_NaCa. (A) Current-voltage (I-V) relationship of current sensitive to external Gd³⁺, obtained by subtracting from control the current after 100 μM Gd³⁺. Inset shows the pulse protocol and representative currents. (B) The effects of Gd³⁺ under conditions in which I_NaCa was abolished by 0Na/0Ca solution. Current in the presence of 0Na/0Ca (minus Gd³⁺) was shown for comparison. Gd³⁺ did not inhibit residual current. Inset shows representative currents.

The sensitivity of I_NaCa to Gd³⁺ was concentration-dependent within the concentration range tested of 2–200 μM. Figure 2A shows the effect of three different [Gd³⁺] on the I-V relationship for I_NaCa (n=6–9 cells for these three concentrations). Similar suppression of both outward and inward current amplitudes was observed at particular [Gd³⁺]. Figure 2B shows the current amplitudes at +60 mV and −100 mV plotted against [Gd³⁺] between 2–200 μM. Half-maximal inhibition was observed at Gd³⁺ concentrations (IC₅₀) of 30.0±4.0 μM at +60 mV (Hill-coefficient, h was 1.04±0.13, n=5–9) and 20.0±2.7 μM at −100 mV (h was 1.13±0.16, n=5–9). These values were not significantly different (P>0.05). The similar IC₅₀ values suggest that the inhibition of I_NaCa by Gd³⁺ was not voltage-dependent. The Hill-coefficient close to 1 suggests that Gd³⁺ exerted its inhibitory effect by at least one molecule of Gd³⁺ binding to one exchanger protein molecule.

Figure 2.

Dose- and voltage-dependence of Gd³⁺ inhibition of I_NaCa. (A) Mean I-V relationship of current amplitudes at different Gd³⁺ concentrations. Gd³⁺ subsequently suppressed I_NaCa. (B) Dose-response curves for the effect of Gd³⁺ on I_NaCa. Currents were measured at +60 mV and −100 mV. Continuous curves were plotted according to the Michaelis-Menten equation: I_Gd³⁺/I_control =1/(1+([Gd³⁺]/IC₅₀)^h). h is Hill coefficient.

In contrast to its effects on I_Ca,L (Lacampagne et al., 1994) or I_Kr (Hongo et al., 1997), Gd³⁺ was slow to exert its full effect on I_NaCa. Application of Gd³⁺ (100 μM) for up to 2 min only partially suppressed I_NaCa while the time for its effect to reach the steady-state was over 6 min. The slow time course of inhibition was not attributable to the perfusion system as (a) Ni²⁺ almost completely suppressed I_NaCa within 12 s, (b) 100 μM Gd³⁺ completely blocked I_Ca,L within only seconds (data not shown). These observations were consistent with previous reports (Lacampagne et al., 1994; Hinde et al., 1999). In addition, reversal of inhibition was rapid on switching to Gd³⁺-free solution. For 100 μM Gd³⁺, the current attained 44.6±5.5% (n=6) of the control value within 12 s of washout. This did not increase with repetitive pulsing; thus, under our conditions reversibility was incomplete. The time-course of Gd³⁺ inhibition of the reverse and forward modes of the exchanger (outward and inward current, respectively) was ascertained by normalizing the mean outward and inward current amplitudes at +60 mV and −100 mV and plotting amplitude relative to control against duration of Gd³⁺ exposure (Figure 3). The decline of both inward and outward I_NaCa was fitted best by a bi-exponential function. The initial decline was slightly faster at +60 mV than at −100 mV; but this difference was not significant (P>0.05). The slower phase of decline was similar at the two potentials.

Figure 3.

Time-dependence of Gd³⁺ inhibition of I_NaCa. Proportion of control I_NaCa after Gd³⁺ was applied at +60 mV and −100 mV plotted against time. Both data sets were fitted with a double exponential function: τ₁=19.2±0.06 s, τ₂=347.4±21.3 s at +60 mV and τ₁=38.6±4.3 s, τ₂=303.0±36.5 s at −100 mV, n=7.

Gd³⁺-inhibition of calcium current has been reported to depend on the extent of channel activation (Biagi et al., 1990). It seemed possible, therefore, that the slow time-course with which Gd³⁺ exerted its steady-state effect might have reflected dependence of the inhibition on repeated activation (use-dependence) of the exchanger. The following approach was used to investigate this. First, the net ‘control' I_NaCa was obtained by measuring the current sensitive to Ni²⁺ (10 mM) application. In seven cells, current was monitored for 5 min during continuous pulsing. In five cells, the control I_NaCa was obtained as above, and then Gd³⁺ (100 μM) applied: cells were exposed to Gd³⁺ for 5 min, but in the presence of 0Na/0Ca solution to immobilize the exchanger and without application of test pulses. (Figure 4, rested in 0Na/0Ca solution). If the block depended on repeated activation of the exchanger, we reasoned that on readmitting external Na and Ca after 5 min (in the presence of Gd³⁺), relatively little block would be expected on the first pulse, but would increase with repeated stimulation. In fact, a large reduction of the current amplitude was observed on the first pulse with this method, to an extent that was not significantly different (P>0.05) from that with continuous pulsing. Similar results were also observed with another approach (to measure the E_rev for I_NaCa for each cell and then hold the membrane potential at this value in the presence of Gd³⁺ (100 μM) for 5 min; the I_NaCa elicited by the first ramp after this period was then measured). These data suggested that the time course of the effect of Gd³⁺ could not be accounted for by slow use-dependence of the inhibition.

Figure 4.

Lack of use-dependence of the effect of Gd³⁺ on I_NaCa. I_NaCa amplitudes at +60 mV under three conditions were compared: Control; 100 μM Gd³⁺ was continuously perfused for 5 min during the time I_NaCa was repetitively activated by continuous pulsing (n=7); Extracellular Na⁺ and Ca²⁺ were replaced by equimolar Li⁺ and Ba²⁺ (0Na/0Ca solution, n=5) during 5 min of Gd³⁺ exposure, then I_NaCa measured when Na and Ca were readmitted. (^****,P<0.00002 with respect to control I_NaCa; ^***P<0.0007 with respect to control I_NaCa).

Discussion

Whilst there has been no previous study that has shown directly inhibition of the sarcolemmal Na⁺-Ca²⁺ exchanger by Gd³⁺, computer-simulations by Hongo et al. (1997) led them to suggest that the blockade of the exchanger was required to reproduce experimental data showing a stretch-induced increase in guinea-pig ventricular [Ca²⁺]_i. Earlier data had suggested that the trivalent cation La³⁺ can inhibit the Na⁺-Ca²⁺ exchanger (Kimura et al., 1986), but direct information regarding Gd³⁺ has not until now been available. Our data show that Gd³⁺ inhibits the cardiac Na⁺-Ca²⁺ exchanger with a potency that compares favourably with other divalent or trivalent inorganic inhibitors: e.g. IC₅₀ for Ni²⁺ of 0.3 mM (Hinde et al., 1999) and inhibition by concentrations >0.25 mM La³⁺ in smooth muscle cells (Shimizu et al., 1997). The Hill co-efficient values near 1 are consistent with observations on other inorganic inhibitors (Blaustein & Lederer, 1999).

The blocking effect of Gd³⁺ on the Na⁺-Ca²⁺ exchanger was found to be voltage-independent, but was dependent on the duration of exposure to Gd³⁺, with block increasing to a steady-state value over a period of minutes after initial application. These observations appear to exclude a surface charge mechanism (which would be expected to show voltage dependence) but are most consistent with Gd³⁺ gradually accumulating at a site that is accessible from the external surface of the cell. On the basis of the experiments shown in Figure 4 it seems reasonable to conclude that the inhibition does not depend on the exchanger being repeatedly activated (in order to facilitate Gd³⁺ binding), as similar levels of block were observed with exchanger immobilization as with repeated activation. Additionally, preliminary data suggested that Gd³⁺ might not be able to readily reach its binding site from the cell interior, because Gd³⁺ in the pipette dialysate (at concentrations up to 5 mM) showed little block of I_NaCa (data not shown). Intriguingly, the trivalent cation La³⁺ has been reported to enter cells (Shimizu et al., 1997; Peeters et al., 1989), and therefore might possibly act at the inside of the plasma membrane.

The demonstration that Gd³⁺ blocks I_NaCa is important both because the exchanger is present in a range of tissue types, and because Gd³⁺ is a widely used experimental tool, particularly in studying stretch-related responses. When interpreting its effects in various experimental systems, potentially diverse actions should be considered (Caldwell et al., 1998). To these, it would appear now necessary to add its potent inhibitory effects on I_NaCa. Further, it is possible that this action of Gd³⁺ may help increase understanding of the range of roles played by the Na⁺-Ca²⁺ exchanger in both physiological and pathological situations.

Abbreviations

E_rev

reversal potential

Gd³⁺

gadolinium

I_Ca,L

L-type calcium current

I_Kr

rapid delayed rectifier current

I_NaCa

Na⁺-Ca²⁺ exchanger current

SAC

stretch-activated channel