Profile of L-type Ca²⁺ current and Na⁺/Ca²⁺ exchange current during cardiac action potential in ventricular myocytes

Abstract

Objective

The L-type Ca²⁺ current (I_Ca,L) and the Na⁺/Ca²⁺ exchange current (I_NCX) are major inward currents that shape the cardiac action potential (AP). Previously, the profile of these currents during AP was determined from voltage-clamp experiments that used Ca²⁺ buffer. In this study, we aimed to obtain direct experimental measurement of these currents during cardiac AP with Ca²⁺ cycling.

Method

A newly developed AP-clamp sequential dissection method was used to record ionic currents in guinea pig ventricular myocytes under a triad of conditions: using the cell’s own AP as the voltage command, using internal and external solutions that mimic the cell’s ionic composition and, importantly, no exogenous Ca²⁺ buffer was used.

Results

The nifedipine-sensitive current (I_NIFE), which is composed of I_Ca,L and I_NCX, revealed hitherto unreported features during AP with Ca²⁺ cycling in the cell. We identified two peaks in the current profile followed by a long residual current extending beyond the AP, coinciding with a residual depolarization. The second peak and the residual current become apparent only when Ca²⁺ is not buffered. Pharmacological dissection of I_NIFE using SEA0400 shows that I_Ca,L is dominant during phase-1&2 whereas I_NCX contributes significantly to the inward current at phase-3&4 of AP.

Conclusion

These data provide the first direct experimental visualization of I_Ca,L and I_NCX during cardiac AP and Ca²⁺ cycle. The residual current reported here can serve as a potential substrate for afterdepolarizations when increased under pathologic conditions.

Introduction

The L-type Ca²⁺ current (I^Ca,L) and the Na⁺/Ca²⁺ exchange current (I_NCX) are two major inward currents that provide the depolarization drive to shape action potential (AP) plateau and repolarization phase in cardiac myocytes. Changes in the magnitude or timing of these currents could cause development of cardiac arrhythmias. Hence, understanding the dynamic properties of these currents during the AP cycle is of great interest. However, the current knowledge on the I_Ca,L and I_NCX during AP has been largely based on model simulations derived from traditional voltage-clamp data. Previous voltage-clamp experiments often used simplified ionic solutions to isolate the currents and used exogenous Ca²⁺ buffer—conditions that differ from physiological milieu. In this study, we directly measured the ionic currents during AP under a triad of conditions mimicking physiological environment: (1) the cell’s own AP was recorded and used as command voltage to directly record the I_Ca,L and I_NCX during AP; (2) the physiological ionic compositions were used in both the internal and the external solutions; and (3) the intracellular Ca²⁺ cycling during AP was preserved by not using exogenous Ca²⁺ buffer in the internal solution.

We recorded the nifedipine-sensitive current (I_NIFE), which is the composite current of I_Ca,L and I_NCX, during AP with Ca²⁺ cycling. Then we used a newly developed AP-clamp Sequential Dissection technique (‘Onion-Peeling’)¹ to separate I_Ca,L and I_NCX using nifedipine and SEA0400. This novel experimental approach allows us, for the first time, to directly visualize the dynamics of I_Ca,L and I_NCX currents during AP with Ca²⁺ cycling in the cell. Our data reveal novel and distinctive features of I_CaL and I_NCX that occur when the cell is most vulnerable to early or delayed afterdepolarizations. Measurement of the currents during the AP with Ca²⁺ cycling under physiological conditions also provide realistic data to aid computational modeling and rational design of effective drug therapies for treating cardiac arrhythmias.

Methods

All laboratory procedures conform to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

Cell isolation

Hartley guinea pigs (male, 3–4 month old, purchased from Charles River Laboratories USA), were first injected with heparin (800u, I.P.) and then anesthetized with nembutal (100 mg/kg, I.P.). After achieving deep anesthesia to suppress spinal cord reflexes, a standard enzymatic technique was used to isolate ventricular myocytes. ²

Electrophysiology

AP-clamp Sequential Dissection experiments were conducted as described in our previous publication¹. Cells were continuously superfused with a modified Tyrode solution (BTy) containing (in mmol/L) NaCl 120, KCl 5, CaCl₂ 2, MgCl₂ 1, HEPES 10, NaHCO₃ 25, Glucose 10, pH 7.3. The pipette solution contained (in mmol/L) K-Aspartate 115, KCl 45, Mg-ATP 3, HEPES 5, cAMP 0.1, pH 7.25. Depending on the experiment, 0, 2 or 10 mM EGTA was added into the pipette solution. Basic experimental steps: (1) Record the steady state AP under I-clamp (I=0) at 1 Hz pacing frequency. (2) Apply this AP as the voltage command onto the same cell under V-clamp at 1 Hz. The net current output, I_BG should be zero. (3) Isolate the current of interest by using its specific blocker to remove it from the net current output, I_drug. (4) The current of interest is then obtained by subtraction: I = I_BG - I_drug. (5) Next, isolate the 2^nd current of interest by applying the 2^nd channel blocker, and then obtained the 2^nd current by subtraction: I₂ = I_drug1 - I_drug2. Repeat (5) to isolate the 3^rd, the 4^th, and more currents by sequentially adding the specific blocker for each channel (Table-1). The currents were recorded after they reached steady state. The inhibitory coefficient of SEA0400 on blocking I_NCX was measured using a standard ramp V-clamp protocol ³. The inhibitory coefficient of SEA0400 on blocking I_CaL was measured using a standard V-clamp protocol ².

Table-1.

Currents (pA/pF)	Ca2+ cycling (0 EGTA)	Ca2+ buffered (EGTA 10 mM)
IK1	Ba2+ 50 µmol/L
Diastolic	0.306 ±0.305	0.281 ±0.215
Peak	1.860 ±0.386	1.723 ±0.541
IKs	Chromanol-293B 10 µmol/L
Diastolic	0.016 ±0.064	−0.007 ±0.138
Peak	0.340 ±0.219	0.374 ±0.208
IKr	E4031 1 µmol/L
Diastolic	0.038 ±0.031	0.004 ±0.110
Peak	0.571 ±0.161	0.724 ±0.184*
INIFE	Nifedipine 10 µmol/L
Diastolic	0.072 ±0.151	0.103 ±0.274
Peak	−1.434 ±0.302	−2.36 ±0.262***

Student’s t-test p<0.05*

p<0.001*** (n=11–17 cells/group)

All experiments were conducted at 21±1°C, except those in Fig.6 were conducted at 36±0.5°C. Chemicals were purchased from SIGMA-ALDIRCH (USA), except SEA0400, 2-[4-[(2,5-difluorophenyl)methoxy]phenoxy]-5-ethoxyaniline was provided by Taisho Pharmaceutical (Japan).

Fig.6.

These experiments were conducted at 36°C. Panel A shows a representative I_NIFE recorded under AP-clamp with Ca²⁺ cycling (n=19 cells). Notice the presence of the second dome and the residual current (A) and the residual depolarization (B). The insert shows the Ca²⁺ transient measured with Fura-2. Panel C shows a representative I_NIFE recorded with 10 mM EGTA (n=9 cells). Notice the absence of the residual current (C) and the residual depolarization (D). Panel E and F show the Mean ±SE and statistical comparison at the features points. (t-test p<0.01**, p<0.001***)

Results

Nifedipine-sensitive current during AP with Ca²⁺ cycling

Ionic currents were recorded in ventricular cardiac myocytes using the Onion-Peeling technique. First, we tested the effect of 0.1% DMSO which is used as solvent of drugs. Application of DMSO resulted in no change in the zero baseline current (Fig.1A). Nifedipine at 10 µM concentration was added into the bath to dissect out I_NIFE, as shown in Fig.1B. I_NIFE displays several distinct features. There is a steep rise of the current at phase-1 of AP, which we named the initial current (“I”) and measured it at 10 ms following the AP upstroke. During phase-2, the current slows down and forms a dome, which we named the dome current and measured it at APD10 (early plateau “E”) and APD20 (dome “D”) (APD# is defined as the AP duration measured at #% repolarization.) During phase-3, the current declines but then resurges to reach a second dome (“S”). Then the current declines again and slowly diminishes during phase-4; we named this feature the residual current (“R”) and measured it at 30 ms after -V_max (the maximum repolarization rate of AP during phase-3). While the initial phase and the dome features were seen before ^4,5, the second dome and the residual current during AP phase-3&4 are novel observations that had not been reported previously.

Fig.1.

Panel A shows a representative current trace recorded with 0.1% (v/v) DMSO (used as solvent of channel blockers). The flat “zero current” indicate a lack of any DMSO effect on membrane currents (n=5 cells). Panel B and C are representative recordings of I_NIFE or I_NISO respectively. Panel D shows the statistical comparison of I_NIFE (n=7) and I_NISO (n=5), demonstrating no significant differences between the currents. Panel E shows the residual depolarization. Panel F shows a positive correlation between the residual depolarization and the residual current measured at 30 ms after −V_max.

To confirm that I_NIFE originates from blocking the L-type Ca²⁺ channel, we used a different blocker, nisoldipine 1 µM to record the nisoldipine-sensitive current (I_NISO) under AP-clamp. I_NISO (Fig.1C, n=8 cells) displays the same distinct features and similar values as I_NIFE (Fig.1D, n=8 cells).

Corresponding to the residual current, we also found a slowing down of the membrane repolarization at the end of AP that extends beyond APD95 (Fig.1E), which we named residual depolarization and measured it as the voltage above the resting potential at 30 ms after -V_max (the same time point for measuring the residual current). The residual depolarization had not been reported in the literature. However, we have consistently found this residual depolarization in both the single cell and the ventricular tissue (using sharp electrode recording of AP, data not shown), confirming its existence in vivo. Furthermore, the residual depolarization shows a strong positive correlation with the residual current (Fig.1F), indicating a connection between the two.

I_NIFE consists of I_Ca,L and I_NCX

I_NIFE is composed of not only I_Ca,L but also I_NCX and other Ca²⁺ sensitive currents. This is because blocking I_Ca,L also eliminates Ca²⁺ entry into the cell to trigger SR Ca²⁺ release, which abolishes the intracellular Ca²⁺ transient and its associated currents such as the inward I_NCX, the Ca²⁺-activated Cl⁻ current, and possibly a fraction of K⁺ currents that is sensitive to Ca²⁺. We conducted all experiments at 21°C to render the Cl⁻ current negligible.^6,7 To determine the Ca²⁺ sensitivity of K⁺ currents, we used the Onion-Peeling method ¹ to record I_Ks, I_Kr, I_K1, and I_NIFE in the same cell. The currents measured with Ca²⁺ cycling versus with EGTA are listed in Table-1. I_K1 and I_Ks did not show any significant difference in the peak current density; peak I_Kr decreased slightly (0.15 A/F). However, the average peak current of I_NIFE was 1.4 A/F with Ca²⁺ cycling and 2.4 A/F with EGTA. Hence, the Ca²⁺ sensitive component in K⁺ currents is much smaller than the magnitude of I_NIFE, which should cause only a slight underestimation of I_NIFE. Therefore, the I_NIFE recorded in our experiment is mainly composed of I_Ca,L and I_NCX.

The Ca²⁺-dependent features of I_NIFE

To separate the I_Ca,L and I_NCX components in I_NIFE, we added 10 mM EGTA to the pipette solution to buffer intracellular Ca²⁺ and minimize the inward I_NCX. A typical recording of I_NIFE with 10 mM EGTA is shown in Fig.2A. Now the I_NIFE profile displays the initial current and the dome current features, but without any discernable second dome and residual current, different from that recorded with Ca²⁺ cycling (Fig.1B). Moreover, the residue depolarization is also eliminated (Fig.2B).

Fig.2.

Representative I_NIFE recorded under AP-clamp with 10 mM EGTA in pipette (n=10 cells). Notice an absence of the second dome, the residual current (A), and the residual depolarization (B). I_NIFE recorded with 2 mM EGTA (C, n=14 cells), the current-voltage relationship (D), and statistical comparison between different Ca²⁺ buffering conditions (E) show increased current density with more EGTA buffering. Panel F demonstrates that EGTA did not affect the voltage at −V_max but significantly reduced the residual depolarization at 30 ms after −V_max. (t-test, p<0.05*; p<0.01**)

When EGTA concentration in the pipette was reduced from 10 mM to 2 mM, I_NIFE displayed the initial current and the dome current but the second dome was absent in all records. Buffering of Ca²⁺ eliminated the second dome and the residual current, but elevated the dome current in concentration-dependent manner (Fig.2A&C). These phenomena are visualized in the I–V relationship (Fig.2D) and confirmed in the statistical comparison of the current amplitudes along the AP time course (Fig.2E). Again, corresponding to a lack of the residual current the residual depolarization was also eliminated (Fig.2F). Note that low EGTA (2 mM) buffering can significantly affect Ca²⁺ and the dynamics of I_NIFE during AP.

Contribution of I_NCX to I_NIFE

The fact that buffering Ca²⁺ can eliminate the second dome and the residual current suggests that I_NCX might contribute to these features. To investigate this we used an I_NCX inhibitor, SEA0400 at 3 µM concentration to record the SEA-sensitive current (I_SEA). As shown in Fig.3A, I_SEA displays a small initial current at AP phase-1; gradually increases during phase-2, peaks at early phase-3 and declines; the current then turns around and reaches the second dome at late phase-3; and then declines to a residual current extending into phase-4. These data show that I_NCX dominates the late phases of AP. After I_SEA had developed and stabilized, we added 10 µM nifedipine into the bath and recorded the I_SEA+NIFE current (Fig.3B, C). Statistical comparison of the currents shows that the contribution of I_NCX to I_SEA+NIFE is relatively small during the early phases of AP but becomes larger in later phases (Fig.3D).

Fig.3.

Panel A shows a representative current recorded with 3 µM SEA0400 (I_SEA). When 10 µM nifedipine was added, I_SEA+NIFE displays similar characteristics (B). Panel C and D show the instant current-voltage relationship of I_SEA and I_SEA+NIFE and statistical comparison of data (n=7 cells, t-test, p<0.05*). Standard V-clamp protocols were used to measure the inhibitory effects of 3 µM SEA. Panel E and F show current traces and the I–V relationship of I_NCX before and after application of SEA. Panel G and H show representative current traces and the I–V relationship of I_Ca,L before and after application of SEA.

Although SEA0400 is the most potent inhibitor of I_NCX currently available, it also partially blocks I_Ca,L. This means that I_SEA is composed of not only I_NCX but also a small portion of I_Ca,L. To separate I_NCX from I_Ca,L, we measured the inhibitory coefficients of 3 µM SEA0400 on I_NCX and I_Ca,L, respectively. The inhibitory effect of SEA0400 on I_NCX is voltage-dependent, about 65% and 51% for the outward I_NCX measured at +30 mV and the inward I_NCX measured at - 75 mV (Fig.3E, F). The inhibitory effect of 3 µM SEA0400 on I_CaL is about 24% at the peak and is also voltage-dependent (Fig.3G, H).

Reconstruction of I_Ca,L and I_NCX during AP

Both I_CaL and I_NCX are heavily dependent on the local Ca²⁺ concentration. However, currently there is no available technique to directly measure the local Ca²⁺ concentrations sensed by the channel/transporter. To circumvent this problem, we took a simple approach to calculate I_CaL and I_NCX from the paired recordings of I_SEA and I_SEA+NIFE (using two equations to solve for two unknowns) by utilizing the differential inhibitory coefficients of SEA versus nifedipine. This simple calculation using paired recordings provides a somewhat direct approach to disentangle I_CaL and I_NCX Given that I_CaL and I_NCX are the predominant currents blocked by nifedipine and SEA0400, the charge conservation gives the following equations:

$$\begin{array}{c}{I}_{\text{SEA}+\text{NIFE}}=I_{\text{CaL}}+\alpha I_{\text{NCX}}\hfill \\ I_{\text{SEA}}=k_{\text{CaL}}(I_{\text{CaL}}+\alpha I_{\text{NCX}})+k_{\text{NCX}}(1-\alpha k_{\text{CaL}})I_{\text{NCX}}\hfill \end{array}$$

where k_CaL and k_NCX are the voltage-dependent inhibitory coefficients of I_CaL and I_NCX measured from the voltage-clamp experiments shown in Fig.3F&H. α is a scaling factor (ranging between 0 and 1 inclusive) that reflects the portion of I_NCX that is affected by changes in I_CaL). Because the local Ca²⁺ concentration sensed by the Na⁺/Ca²⁺ exchanger is unknown, we therefore solved the equations for α =0 and 1 to calculate the boundary conditions and estimate the currents. Fig.4 show I_NCX and I_CaL computed from paired recordings of I_SEA and I_SEA+NIFE during AP with Ca²⁺ cycling (Column A) or with EGTA (Column B). The current traces are shown in the upper row. Notice that I_SEA is an inward current due to the forward mode INCX when Ca²⁺ is cycling, but turns into an outward current due to the reverse mode I_NCX when Ca²⁺ is buffered with 10 mM EGTA. The I_Ca,L and I_NCX calculated from I_SEA and I_SEA+NIFE are shown in Fig.4 (middle and lower row). The blue (α =1) and red (α =0) traces demarcate the upper and the lower boundaries of the currents, respectively. The actual current should fall within the boundaries. Note that I_NCX is bounded within a narrow range; hence the profile of I_NCX during AP is clearly defined. A probable profile of the current is highlighted with black line (calculated as an intermediate state with α=0.8).

Fig.4.

Column A and B show I_SEA and I_SEA+NIFE recorded in the absence and presence of EGTA, respectively. EGTA buffering of Ca²⁺ reversed I_SEA and amplified I_SEA+NIFE. Mathematical reconstruction of I_Ca,L and I_NCX assuming zero coupling (α=0, red line) or linear coupling (α=1, blue line) between I_Ca,L and I_NCX mark the boundaries of the actual currents. Black line shows an intermediate state (α=0.8), highlighting a probable profile of the currents.

The magnitude of I_CaL dome current is about 0.3 A/F when Ca²⁺ is normally cycling (Column A, middle panel), but becomes much greater to about 2 A/F when Ca²⁺ is buffered (Column B, middle panel). This difference reflects a profound influence of the Ca²⁺-dependent inactivation of I_Ca,L. Note also that buffering Ca²⁺ completely eliminated the second dome and the residual current. When Ca²⁺ is cycling, I^NCX is briefly outward at the beginning of AP but soon becomes an inward current (Column A, lower panel), and reaches a peak of about 0.2 A/F. In contrast, when Ca²⁺ is buffered, I_NCX is always outward during AP (Column B, lower panel).

Effects of SEA0400

It has been proposed that inhibiting I_NCX might provide an effective therapy for some forms of arrhythmias. We predicted that using SEA0400 to treat the cell would reduce the dome current and the residual current. To test this, we pretreated cells with 3 µM SEA0400 and recorded the remaining I_NIFE under AP-clamp. As shown in Fig.5A, SEA0400 treatment shortened APD, in consistent with its effect on reducing the dome and the residual current.

Fig.5.

Using 3 µM SEA0400 to treat cells shortened AP duration (A) and reduced I_NIFE (B). The residual current (C) and the residual depolarization (D) were significantly reduced (n=7 cells, t-test, p<0.05*), demonstrating a predominant influence of I_NCX on these features. Panel E shows a universal correlation (R=0.917) between the magnitude of residual current and residual depolarization under all experimental conditions tested.

A representative trace of the I_NIFE in the SEA0400 pretreated cell is shown in Fig.5B. Recall that in the control cell (Fig.1B) the dome is larger than the initial current, here in the SEA0400 pretreated cell the dome becomes smaller than the initial current, indicating a portion of the dome current is contributed by the inward I_NCX. The residual current is significantly smaller after SEA0400 treatment (Fig.5C), in consistent with the I_NCX contribution. Meanwhile the residual depolarization is also smaller (Fig.5D). In fact, a positive correlation between the residual current and the residual depolarization exists universally for all the experimental conditions tested (Fig.5E), indicating that the residual current is responsible for the residual depolarization.

I_NIFE at body temperature

To establish the physiological relevance of our new findings, we also conducted AP-clamp experiments at body temperature. The I_NIFE at 36°C displays essentially the same features (Fig.6A). The Second dome and the residual currents are prominent, and the residual depolarization is also present (Fig.6B). The insert shows the Ca²⁺ transient measured with Fura-2. Using EGTA to buffer Ca²⁺ eliminates the residual current (Fig.6C) and the residual depolarization (Fig.6D). Statistical comparison show that with Ca²⁺ cycling, the dome currents is significant smaller (Fig.6E) due to Ca²⁺ -dependent inactivation of I_Ca,L (CDI), but the residual current and the residual depolarization are significantly larger (Fig.6E,F) due to the inward I_NCX.

Discussion

Here we present experimental measurements of I_Ca,L and I_NCX currents during the action potential with Ca²⁺ cycling in ventricular myocytes. This was enabled by a triad of conditions: (1) using the cell’s own AP as the voltage command; (2) using the internal and external solutions that mimic the physiological milieu; and (3) Ca²⁺ cycling during AP was preserved by not using exogenous Ca²⁺ buffer. To the best of our knowledge, it is the first time this triad of conditions has been used to record the ionic currents during cardiac AP with Ca²⁺ cycling.

I_NIFE was obtained by blocking L-type Ca²⁺ current under AP-clamp. However, blocking Ca²⁺ current also affects other Ca²⁺ dependent currents including I_NCX, Ca²⁺-activated Cl⁻ current, and the Ca²⁺ sensitive components in K⁺ currents. The Cl⁻ current was rendered negligible by conducting experiments at 21°C.^6,7 We also found no significant contributions of K⁺ currents. Hence, I_NIFE mainly consists of I_Ca,L and I_NCX. The profile of I_NIFE during AP exhibits distinct features: an initial current at phase-1, a dome at phase-2, a resurging second dome at phase-3 and a residual current at phase-4. The second dome and the residual current had not been reported before; these novel features become apparent only when Ca²⁺ is not buffered. We also detected a hitherto unreported residual depolarization at the end of AP which can be attributed to the residual current. Furthermore, we found that I_Ca,L contributes predominantly to the initial and the dome current whereas I_NCX contributes significantly to the second dome and the residual current. The timing and magnitude of these currents suggest their relative roles in shaping the AP profile and generating early or delayed afterdepolarizations.

The dynamic profile of I_Ca,L during cardiac AP and its contribution to afterdepolarizations has been under active investigation. Previous studies used variants of the AP-clamp technique and reported the initial current and the dome current features, but never before reported the second dome and the residual current features. This discrepancy can be explained by the experimental conditions used in those studies that differed from the triad of conditions used in our experiments. First, in earlier studies the AP waveform used as voltage command was either a representative AP ⁵ or a reconstructed AP ⁸, which does not, in general, match each cell’s unique AP. Hence the current profile would be distorted from its natural state in the cell. Second, earlier experiments used non-physiological solutions (substituting Na⁺ and K⁺), but ion species and concentration are known to alter the Ca²⁺ channel kinetics. ^9,10 Third, most earlier experiments used exogenous Ca²⁺ buffer, and the I_Ca,L profile reported in those studies ^11,8,5 are similar to the I_NIFE profile recorded with EGTA in our study (Fig.2A&C). The second dome and the residual current in I_NIFE are novel features revealed only under the triad of conditions used in our AP-clamp experiments.

The dynamic profile of I_Ca,L during AP is determined by an interplay between the changing driving force during AP and CDI. Importantly, our data show a significant modulation of I_Ca,L during AP by CDI, seen as the I_NIFE dome current increases with EGTA in a dose-dependent manner (Fig.2E). The average dome current is 1.4 pA/pF when Ca²⁺ is cycling, and increases to 2.4 pA/pF when Ca²⁺ is buffered with10 mM EGTA. It is noteworthy that previous voltage-clamp and AP-clamp studies (using Ca²⁺ buffer) reported a peak I_Ca,L between 2–10 pA/pF ^11,5,4; quantitative models based on those data also had a peak I_Ca,L in that range ¹². Our data show a much smaller I_Ca,L magnitude in the cell measured under AP-clamp with Ca²⁺ cycling, and provide realistic experimental data for future modeling considerations.

Previously, a lack of selective blocker prevented direct measurement of I_NCX ¹³. In this study we used SEA0400 paired with nifedipine to isolate the I_NCX during AP with Ca²⁺ cycling. Our data reveal that I_NCX presents an outward current at the beginning of AP, which turns into a small inward current at early phase-2; the inward I_NCX increases and contributes significantly to the I_NIFE second dome and the residual current at phase-3&4. In comparison to earlier studies, Grantham and Cannell ⁵ had calculated an I_NCX profile that showed a large outward current throughout AP phase-1&2, then turned into an inward current at phase-3, and then peaked at the end of phase-3. Weber et al. ¹⁴ reconstructed the I_NCX during AP by correlating the Ca²⁺ transient with the steady-state I_NCX measurements at various Ca²⁺ concentrations. They showed that, in rabbit ventricular myocytes, I_NCX was a rapid outward current at phase-1, turned to inward at early phase-2, reached a peak at the end of phase-3, and then slowly declined during phase-4. The I_NCX profile in Fig.4 (Column A, lower panel) shows similar features as their result, but differs in the time course and the magnitude of current. Given the importance of I_NCX in shaping AP, our AP-clamp measurement of I_NCX in physiological milieu provides important data for developing more accurate quantitative models.

Increased I_Ca,L or I_NCX had been found to induce afterdepolarizations ^15,16. Inhibiting I_Ca,L with Ca²⁺ channel blocker effectively eliminated early afterdepolarizations. ^{15, 17, 18} Inhibiting I_NCX with SEA0400 also effectively reduced early and delayed afterdepolarizations. ^3,13 Hence, I_Ca,L and I_NCX have been proposed as therapeutic targets for treating cardiac arrhythmias (see review by Sipido et al. ¹³). The relative contributions of I_Ca,L and I_NCX to the depolarization drive are resolved in the current study. As we observed, I_Ca,L is the dominant inward current during AP phase-1&2, whereas I_NCX is the dominant inward current during phase-3&4. Both I_Ca,L and I_NCX contribute to the dome current, the second dome, and the residual current. Given the intertwined nature of I_Ca,L and I_NCX, it is the combination of these two currents, not just one or the other, determines the total depolarization drive that shapes AP profile and afterdepolarizations. ¹⁹ Importantly, the newly observed second dome and the residual current occurring at phase-3&4 have the right timing and the shape to generate afterdepolarizations if not counterbalanced by repolarization currents. Our data clearly show that the second dome coincides with the vulnerable phase of AP where early afterdepolarizations may develop, while delayed afterdepolarizations may occur at where we observed the residual current. The current study provides useful and physiologically relevant data to aid quantitative modeling of cardiac AP and rational design of anti-arrhythmia drug therapies.