INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MYOCARDIAL STRETCH CONTRIBUTES to arrhythmogenesis by modulating cellular electrical activity (6, 13, 27, 31). Stretch induces transient depolarizations (SIDs) from phase 4 and premature beats, increases dispersion of repolarization, and reduces the threshold for fibrillation. Mechanoelectrical feedback is thought to result, at least in part, from activation of poorly selective stretch-activated cation channels (SACs) (21, 39). One argument favoring this conclusion is the effect of Gd³⁺, a trivalent lanthanide that blocks cation SACs (16, 41), on the response to stretch. Gd³⁺ suppresses SIDs and premature beats that accompany increased volume in isolated left ventricle (18, 34), aortic cross-clamping (36), or increasing intra-atrial pressure (37). Shortened action potential duration and increased dispersion of repolarization associated with stretch also are prevented by Gd³⁺ pretreatment (36). In addition, a volume-sensitive cation current activated by osmotic swelling (10) and in a tachycardia-induced dilated cardiomyopathy (11) is inhibited by Gd³⁺.

The utility of Gd³⁺ in establishing a role for SACs in stretch-induced arrhythmias depends on its specificity. Perhaps because Gd³⁺ is a lanthanide, concerns regarding specificity have focused on block of L- and T-type Ca²⁺ channels in heart (22, 24, 29) and other tissues (4, 23, 26). Besides being a potent Ca²⁺ channel blocker, La³⁺ also modulates Na⁺ current (I_Na) in neurons (1, 40), pituitary GH3 cells (2, 3), and cardiac myocytes (17, 32) in a complex manner. Less information is available about the effects of other lanthanides on I_Na. Gd³⁺ reduces I_Na and slows gating in myelinated Xenopus laevis axons (12), but its actions on the cardiac Na⁺ channel isoform are essentially unknown (7). Studies in heart are important, because block of I_Na by cations is isoform dependent, as shown for group 2B cations Cd²⁺ and Zn²⁺ (14).

This study examined whether block of cardiac Na⁺ channels by Gd³⁺ might contribute to its antiarrhythmic efficacy on stretch-induced ectopic activity. Gd³⁺ shifted activation and availability of Na⁺ conductance to more positive potentials, reduced maximum Na⁺ conductance (g_max), and slowed activation and inactivation kinetics. These effects made Gd³⁺ a potent blocker of cardiac I_Na, especially near threshold potential. Nevertheless, in some situations, Gd³⁺ paradoxically increased I_Na, an effect that could be explained by the positive shift of availability.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Myocyte preparation and solutions. Rabbit ventricular myocytes were enzymatically dissociated (10, 25). Briefly, New Zealand White rabbits of either gender (2-3 kg) were anesthetized, and their hearts were quickly removed and placed in oxygenated Tyrode solution. Hearts were mounted on a Langendorff column and perfused with 37°C oxygenated Tyrode solution containing (mM) 140 NaCl, 5.4 KCl, 1 MgCl₂, 2 CaCl₂, 0.33 NaH₂PO₄, 10 glucose, and 10 HEPES (pH adjusted to 7.4 with NaOH). Then, after perfusion with Ca²⁺-free Tyrode solution for 5-10 min, hearts were digested with solution containing 0.5 mg/ml collagenase (type II, Worthington) and 1 mg/ml BSA (Sigma Chemical). Isolated myocytes were stored in high-K⁺ media containing (mM) 10 KCl, 10 KH₂PO₄, 120 potassium glutamate, 10 taurine, 1.0 MgSO₄, 20 glucose, and 0.5 EGTA (pH adjusted to 7.2 with KOH) and used within 8 h.

Electrophysiology and data analysis. I_Na was recorded by the whole cell patch-clamp technique (Axopatch 200A, Axon Instruments). Data were acquired at 50 kHz, and command pulses were generated by a Digidata 1200B (Axon) controlled by pClamp 7 (Axon). Recordings were low-pass filtered at 5 kHz with an eight-pole Bessel filter and stored on hard disk.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of Gd³⁺ on I_Na. Families of capacity- and leak-corrected I_Na elicited with 30-ms steps from 120 mV to between 80 and +40 mV are shown in Fig. 1, A and B. Inward and outward currents were substantially reduced by 50 µM Gd³⁺. The time course of block of I_Na by Gd³⁺ at 35 mV and recovery on washout are illustrated in Fig. 1C. Block reached a steady state after ~10 min, but recovery was biphasic and incomplete during the 10-min washout period. Incomplete recovery from block by Gd³⁺ also was reported for L-type Ca²⁺ channels (24). Block of peak I_Na was associated with slowing of activation and inactivation kinetics. In superimposed traces (Fig. 1D), slowing of kinetics is reflected as prolonged time to peak and crossing of current traces during inactivation.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Gd3+ (50 µM) blocks Na+ current (INa) in ventricle. A: control; 30-ms steps from 120 mV to between 80 and +40 mV at 0.1 Hz. B: Gd3+; same cell as A. C: time course of block at 35 mV; washout was incomplete. D: INa at times (a-d) indicated in C. Gd3+ prolonged time to peak inward current and slowed inactivation.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   Dose dependence of block of INa. INa with Gd3+ normalized by control (INa-Gd3+/INa-Cont), test potential (TP) = 35 mV, holding potential = 120 mV. Data fitted cell by cell to the following equation: INa-Gd3+/INa-Cont = 1/[1 + (C/IC50)b], where IC50 is concentration giving 50% inhibition, C is Gd3+ concentration, and b is Hill coefficient. IC50 was 47.8 ± 4.8 µM, and b was 1.27 ± 0.05 (n = 27 cells, 6 concentrations each). Fit to plotted mean data (solid line): IC50 = 40.6 µM; b = 1.07.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Voltage dependence of block. Currents recorded as in Fig. 1. A: control. B: 100 µM Gd3+; same cell as in A. C: control current-voltage (I-V) relationships and with 50 and 100 µM Gd3+ (n = 7). D: fraction of current remaining (INa-Gd3+/INa-Control) as function of voltage, calculated from mean data in C.

Voltage dependence of conductance and availability. The voltage dependence of the conductance activation variable (g/g_max) was determined from I-V relationships for each cell (Fig. 4A) and was fitted to the Boltzmann equation to obtain voltage for half-activation (V_0.5) and slope factor (S). The voltage dependence of availability (I/I_max) was determined as illustrated in Fig. 4B and also fitted to the Boltzmann equation. Figure 4C shows that 50 µM Gd³⁺ shifted the midpoint for conductance and availability of I_Na to more positive potentials. V_0.5 for activation shifted 7.9 mV, from 43.1 ± 1.5 mV in control to 35.2 ± 1.3 mV in Gd³⁺ (n = 7, P < 0.05). The shift of the availability curve was slightly less, 5.7 mV, from 82.8 ± 1.4 mV in control to 77.1 ± 1.2 mV in Gd³⁺ (n = 7, P < 0.05). In contrast, S values were not significantly altered. Values of S for conductance were 7.1 ± 0.5 and 8.9 ± 0.6 mV (n = 7, not significant) and for availability were 5.6 ± 0.6 and 5.3 ± 0.5 mV (n = 7, not significant) in control and Gd³⁺, respectively.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Positive shift of activation and availability. A: I-V curves before and after 50 µM Gd3+ used to calculate activation (g/gmax, where g is conductance). Protocol as in Fig. 1. B: protocol and representative recordings used to assess availability (I/Imax). Currents (I) at 30 mV after 1-s conditioning pulses to between 140 and 50 mV were normalized by maximum current (Imax). C: voltage dependence of activation and availability in control and with 50 µM Gd3+ (n = 7). I/Imax and g/gmax were fitted to Boltzmann distribution: y = 1/{1 + exp[(Vm  V0.5)/S]}, where Vm is membrane potential, V0.5 is the midpoint, and S is slope. For activation, V0.5 and S were 43.1 ± 1.5 mV and 7.1 ± 0.5 mV in control and 35.2 ± 1.3 mV and 8.9 ± 0.6 mV in Gd3+. For availability, V0.5 and S were 82.8 ± 1.4 mV and 5.6 ± 0.6 mV for control and 77.1 ± 1.2 and 5.3 ± 0.5 mV in Gd3+.

Kinetics of activation and inactivation. Kinetics of activation and inactivation of inward I_Na were assessed as illustrated in Fig. 5A. Currents during 30-ms steps from 120 to 35 mV and superimposed monoexponential fits of activation and biexponential fits of inactivation are shown. Raw traces (Fig. 5A) and mean data show that Gd³⁺ significantly slowed the time constants for inactivation (_h1 and _h2; Fig. 5B) and activation (_m; Fig. 5D) in a dose-dependent manner (n = 7, P < 0.01). Slowing of gating was relatively less at more positive potentials, however. In addition, the fraction of I_Na that underwent rapid inactivation was reduced by Gd³⁺ at 50 (P < 0.05) and 100 µM (P < 0.01). Figure 5C shows the voltage dependence of the fraction of I_Na that rapidly inactivates (A₁/A_total, where A₁ is the amplitude of the fast component of inactivation and A_total is the sum of the fast and slow components obtained from biexponential fits to the decay of I_Na). At 30 mV, for example, A₁/A_total decreased from 0.86 ± 0.03 in control to 0.78 ± 0.05 and 0.34 ± 0.08 in 50 and 100 µM Gd³⁺.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Gd3+ slowed kinetics of activation and inactivation. A: INa on 30-ms steps to 35 mV from 120 mV in control and 50 µM Gd3+. Activation (m) and inactivation (h1 and h2) time constants were obtained from superimposed exponential fits. Left, bar at 0 pA/pF; right, steady-state current. B: h1 and h2 in control and 50 and 100 µM Gd3+ (n = 7) plotted as a function of test potential. C: fractional amplitude of fast component (A1/Atotal) plotted as function of test potential. Gd3+ significantly decreased A1/Atotal in a dose-dependent manner. D: m in control and 50 and 100 µM Gd3+ (n = 7) plotted as a function of test potential.

Kinetics of recovery from inactivation. Recovery of I_Na from inactivation was studied with a paired-pulse protocol. Superimposed currents and the time course of recovery are illustrated in Fig. 6. I_Na recovery was complete and well fitted by monoexponential functions with time constants of 5.9 ± 0.3 ms in control and 6.4 ± 0.4 ms in 50 µM Gd³⁺ (n = 8, not significant). This indicates that Gd³⁺ did not affect recovery of I_Na from inactivation at hyperpolarized potentials.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Gd3+ did not affect recovery from inactivation. Kinetics of recovery assessed with paired 100-ms pulses (P1 and P2) to 30 mV from 120 mV at 0.1 Hz with varying P1-P2 intervals (inset). A: typical currents are superimposed. B: P2 current normalized by P1 current and plotted vs. P1-P2 interval. Recovery was fit with monoexponential functions; time constants were 5.9 ± 0.3 ms for control and 6.4 ± 0.4 ms for 50 µM Gd3+ (n = 8).

Holding potential-dependent effects of Gd³⁺. The positive shifts of the activation and availability curves (Fig. 4C) suggested that the effects of Gd³⁺ on I_Na would depend on resting potential or, under voltage clamp, on holding potential. Figure 7 compares families of currents and I-V relationships obtained with holding potentials of 90 mV (Fig. 7, A-C) and 75 mV (Fig. 7, D-F) under control conditions and in 50 µM Gd³⁺. These two holding potentials approximate the range of resting potential under physiological conditions. When membrane potential was held at 90 mV, Gd³⁺ reduced inward and outward I_Na, as previously shown when holding potential was set to 120 mV (Fig. 4). At a holding potential of 75 mV, however, Gd³⁺ increased I_Na at all potentials positive to 40 mV and produced a crossover of the control and Gd³⁺ I-V relationships (Fig. 7F). Similar results were obtained in five myocytes studied at both holding potentials. At a holding potential of 90 mV, 50 µM Gd³⁺ decreased maximum inward I_Na from 13.7 ± 1.5 to 9.1 ± 1.1 pA/pF (n = 5, P < 0.01) at 30 mV, whereas Gd³⁺ increased I_Na from 3.1 ± 0.5 to 4.3 ± 0.7 pA/pF (n = 5, P < 0.01) at a holding potential of 75 mV. As expected, however, reducing Na⁺ channel availability by holding at 90 or 75 mV, rather than 120 mV (Fig. 4), did not significantly affect the midpoint or slope factor of the activation curves in the presence or absence of Gd³⁺ (data not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Effect of physiological holding potential. INa with 90 and 75 mV holding potential (note 4-fold change in current scale) in control (A and D) and with 50 µM Gd3+ (B and E). I-V relationship (C and F) in control and with Gd3+. Gd3+ shifted threshold for INa activation positive in both cases, as previously shown with holding potential of 120 mV (Figs. 3 and 4). Gd3+ decreased INa at all test potentials with holding potential of 90 mV, as previously shown for 120 mV, but paradoxically increased INa positive to 40 mV with holding potential of 75 mV. Each cell (n = 5) was studied with and without Gd3+ at both holding potentials.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Gd³⁺ has complex effects on I_Na in rabbit ventricular myocytes and produced 1) a positive shift of activation and availability, 2) slowing of activation and inactivation kinetics, and 3) reduction of g_max. The positive shift of activation and decreased g_max combined to make Gd³⁺ a potent inhibitor of I_Na, especially near threshold potential.

Typically, 10-100 µM Gd³⁺ is used to block SACs and stretch-induced arrhythmias. At these concentrations, a significant reduction of I_Na and a positive shift of threshold potential is expected. Gd³⁺ was a more potent blocker of I_Na near threshold potential than at more positive voltages; IC₅₀ was estimated as 15 µM at 55 mV and 48 µM at 35 mV. Voltage dependence of block arose in large part because Gd³⁺ shifted the voltage dependence of activation. Consequently, by virtue of its effects on Na⁺ channels, Gd³⁺ should be particularly effective at blocking ectopic activity arising from SIDs near threshold potential, and block of I_Na must be considered a possible antiarrhythmic mechanism before it can be concluded that suppression of stretch-induced electrical activity by Gd³⁺ implicates poorly selective cation SACs in the process. Because the response to stretch was not studied in the present experiments, conclusions regarding the relative importance of the proposed antiarrhythmic actions of Gd³⁺ (i.e., block of SACs or block of I_Na) cannot be drawn.

Gd³⁺ paradoxically increased I_Na at voltages positive to 40 mV when holding potential was reduced to 75 mV, whereas Gd³⁺ inhibited I_Na at all potentials when holding potential was 90 or 120 mV. Nevertheless, threshold for activation of I_Na was shifted positive at all holding potentials. This phenomenon arises when availability initially is low, because the Gd³⁺-induced shift enhances availability. Gd³⁺ doubled availability at 75 mV, from 0.20 to 0.41, and this outweighed the other actions of Gd³⁺ that reduce I_Na. A similar paradoxical stimulation of I_Na is likely with many interventions (e.g., divalent and other trivalent cations) that shift availability positive but otherwise block I_Na. Because analysis of I_Na usually is undertaken from holding potentials that provide full availability, this paradoxical stimulation is not well known. It may be functionally relevant in situations with reduced resting potential.

To our knowledge, the only previous study of the effect of Gd³⁺ on cardiac I_Na was by Bustamente (7), who studied three human atrial cells and reported that 100 µM Gd³⁺ decreased I_Na, slowed kinetics, and reduced membrane capacitance by 37%. The large reduction in capacitance over time is a confounding factor, however. We and others (12, 22, 29) did not note that Gd³⁺ decreased capacitance.

Gd³⁺ appears more potent than La³⁺ for shifting activation and, thereby, blocking cardiac I_Na near threshold. Gd³⁺ at 50 µM shifted the activation midpoint by 7.9 mV, whereas 95 µM La³⁺ shifted activation only 5 mV in Purkinje cells (17). The effects of Gd³⁺ on cardiac I_Na also can be compared with those reported in Xenopus node of Ranvier (12). The IC₅₀ for block of I_Na at 10 mV, 74 µM, was similar to the 64 µM estimated here; 60 µM Gd³⁺ shifted the midpoints of activation and availability by 6 and 10 mV, respectively, in nerve, whereas 50 µM Gd³⁺ shifted these parameters by 7.9 and 5.7 mV in heart, and reduction of g_max in heart by 60% was approximately twofold greater than the decrease in peak Na⁺ permeability in nerve. Thus Gd³⁺ is a potent blocker of I_Na in mammalian cardiac muscle as well as amphibian peripheral nerve.

Mechanisms of block. Divalent and trivalent cations cause a positive shift in the voltage dependence of ion channel gating by screening or binding to negatively charged sites on the membrane's extracellular face, thereby altering surface potential (_o) (19). The Grahame equation relates surface charge density (, in e nm²) to _o (in mV)

(1)

where C_k is concentration (in mol/l) of species k of valence z_k, R is the gas constant, T is absolute temperature, and F is Faraday's constant (19). If it is assumed that  is 0.72 e nm² in control bathing solution (12, 17) and that Gd³⁺ screens but does not bind (Gouy-Chapman-Stern model), 50 and 100 µM Gd³⁺ are expected to reduce _o from 58.9 mV to 58.3 and 57.7 mV. The predicted positive shifts, 0.6 and 1.2 mV, are much less than the observed shifts of activation and availability (Fig. 4) and of the time constants for these processes (Fig. 5). This suggests that Gd³⁺ also must bind to charged sites within a Debye length of the voltage sensor. Binding of Gd³⁺ would reduce effective surface charge density and, thereby, _o (Gouy-Chapman-Stern model), as previously argued to explain shifts in Na⁺ channel activation and availability with divalent and trivalent cations (19), including Gd³⁺ in nerve (12) and La³⁺ in heart (17). Estimation of the dissociation constant for the Gd³⁺-binding site interaction from the observed positive shifts is highly sensitive to assumptions regarding surface charge density and is not further addressed here.

(2)

Utility of Gd³⁺ as a selective blocker of SACs. After Yang and Sachs (41) reported block of SACs by Gd³⁺, Gd³⁺ was widely adopted as a tool to study mechanogated channels (16, 21, 29, 37) and stretch-induced arrhythmias (18, 34, 36). As recognized in the original report (41) and subsequent publications (9, 16, 22, 29), care must be taken in interpreting experiments, because Gd³⁺ can modulate a number of channels. Besides its effects on Na⁺ channels in heart and nerve (12), Gd³⁺ also is a potent blocker of Ca²⁺ channels in cardiac (22, 24, 29), neuronal (23, 26), and pituitary cells (4). Gd³⁺ also inhibits certain K⁺ and Cl channels. In guinea pig ventricular myocytes, Gd³⁺ blocks the rapid component of the delayed rectifier, I_Kr, but not the slow component, I_Ks, or the inward rectifier, I_K1 (20, 28). These effects are similar to those of La³⁺, which inhibits I_Kr and shifts the activation I_Ks without reducing its g_max (30). Delayed rectifier K⁺ current in nerve also is blocked by Gd³⁺ (12). Swelling-induced, Ca²⁺-dependent activation of a Cl current in embryonic cultured chick heart myocytes is inhibited by Gd³⁺ (15), as is activation of endogenous Ca²⁺-activated Cl current in Xenopus oocytes (38). On the other hand, swelling-activated Cl current in rabbit ventricle is not affected by Gd³⁺ (10). Finally, Gd³⁺ blocks a lysolipid-induced time-independent inward current carried largely by Na⁺ in ventricle (8) and decreases leakage current in nerve (12).

Conclusions. The present observations provide strong evidence that Gd³⁺ inhibits I_Na in mammalian cardiac myocytes at concentrations used to block SACs. The positive shift in the voltage dependence of activation and reduction in g_max combine to make Gd³⁺ a potent inhibitor of I_Na near threshold potential. Therefore, suppression of stretch-induced arrhythmias by Gd³⁺ may not be sufficient evidence to conclude that SAC activation is responsible for altered cardiac rhythm.