L-Type Calcium Channel Inhibition Contributes to the Proarrhythmic Effects of Aconitine in Human Cardiomyocytes

Abstract

Aconitine (ACO) is well-known for causing lethal ventricular tachyarrhythmias. While cardiac Na+ channel opening during repolarization has long been documented in animal cardiac myocytes, the cellular effects and mechanism of ACO in human remain unexplored. This study aimed to assess the proarrhythmic effects of ACO in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs). ACO concentration-dependently (0.3 ~ 3.0 μM) shortened the action potentials (AP) durations (APD) in ventricular-like hiPSC-CMs by > 40% and induced delayed after-depolarization. Laser-scanning confocal calcium imaging analysis showed that ACO decreased the duration and amplitude of [Ca2+]i transients and increased in the beating frequencies by over 60%. Moreover, ACO was found to markedly reduce the L-type calcium channel (LTCC) currents (ICa,L) in hiPSC-CMs associated with a positive-shift of activation and a negative shift of inactivation. ACO failed to alter the peak and late Na+ currents (INa) in hiPSC-CMs while it drastically increased the late INa in Guinea-pig ventricular myocytes associated with enhanced activation/delayed inactivation of INa at -55 mV~ -85 mV. Further, the effects of ACO on ICa,L, INa and the rapid delayed rectifier potassium current (Ikr) were validated in heterologous expression systems by automated voltage-clamping assays and a moderate suppression of Ikr was observed in addition to concentration-dependent ICa,L inhibition. Lastly, increased beating frequency, decreased Ca2+ wave and shortened field potential duration were recorded from hiPSC-CMs by microelectrode arrays assay. In summary, our data demonstrated that LTCC inhibition could play a main role in the proarrhythmic action of ACO in human cardiomyocytes.

About the Authors:

Jianjun Wu

Affiliation: National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore

Xiangchong Wang

Affiliation: Department of Pharmacology, Hebei Medical University, Shijiazhuang, Hebei, China

Ying Ying Chung

Affiliation: National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore

Cai Hong Koh

Affiliation: National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore

Zhenfeng Liu

Affiliation: National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore

Huicai Guo

Affiliation: Department of Toxicology, Hebei Medical University, Shijiazhuang, Hebei, China

Qiang Yuan

Affiliation: Neuroscience & Behavioral Disorders Program, Duke-NUS Medical School Singapore, Singapore

Chuan Wang

Affiliation: Department of Pharmacology, Hebei Medical University, Shijiazhuang, Hebei, China

Suwen Su

* E-mail: [email protected] (HW); [email protected] (SS)

Affiliation: Department of Pharmacology, Hebei Medical University, Shijiazhuang, Hebei, China

Heming Wei

* E-mail: [email protected] (HW); [email protected] (SS)

Affiliations National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore, Cardiovascular & Metabolic Disorders Program, Duke-NUS Medical School Singapore, Singapore

Introduction

Aconitum is a well-known medical herb which has been used world widely for over 2000 years. However, aconitine (ACO), the main effective ingredient of aconitum and a highly toxic diterpenoid alkaloid, has long been associated with severe cardiovascular toxicities including tachyarrhythmia and hypotension that causes a high mortality in patients [1,2].

Previous studies indicate that ACO is capable of inducing ventricular tachycardia (VT) and ventricular fibrillation (VF) in mouse [3], rat [4], Guinea-pig [5] and rabbits [6]. In isolated sheep heart Purkinje fibers, ACO has been shown to act as an cardiac Na+ channel agonist that opens the Na+ channels during the depolarization/repolarization phase of an action potentials (AP), leading to a delayed repolarization and early after-depolarization [7]. Similar effects of ACO were found in isolated ventricular myocytes of mice [8] and rats [9] and potentially in Guinea-pigs as well [10].

In addition, ACO-induced L-type Calcium channel (LTCC) inhibition has been observed in neonatal [11] and isolated adult [12] rat cardiac myocytes. Accumulated intracellular Na+ has been recognized for activating the reverse-mode of Na+-Ca2+ exchanger (NCX) and increasing cytosolic [Ca2+]i, which can trigger intercellular [Ca2+]i concentration dependent inactivation (CDI) of the LTCC [13,14]. It remains unclear if LTCC inhibition by ACO is due to accumulation of [Na+]i.

There are other possible mechanisms involved in ACO-induced toxicity. ACO has also been found capable of blocking the Kv1.5 channels that mediate the outward ultra-rapidly activating delayed rectifier K+ current (IKur) in neonatal rat ventricular myocytes [15,16] while studies also demonstrated that ACO inhibited hERG in Xenopus laevis oocytes [16] and HEK293 cells [17].

Despite many intoxication incidences occurred each year, the proarrhythmic effects of ACO in human and the underlying mechanism remain unexplored due to a lack of human cardiomyocyte models. The recent scientific breakthrough has enabled the generation of human induced pluripotent stem cells (hiPSCs)-derived cardiomyocytes (hiPSC-CMs) which share important structural and functional similarities with native human cardiac myocytes. To date, hiPSC-CMs have been successfully adopted for modeling various cardiac diseases [18,19] and for drug testing [20,21].

We hypothesize that ACO-induced arrhythmia in human may involve an alternative mechanism different from that observed in animals due to a marked inter-species variation in cardiac electrophysiology [22,23].

In the present studies, we showed that ACO is able to block the ICa,L and calcium-induced calcium release (CICR) in hiPSC-CMs without altering INa. Such effects could contribute to the shortened repolarization period characterized by shortened AP duration (APD) and field potential duration (FPDc).

Materials and Methods

Chemicals and drugs

ACO, Bay K-8644, and nifedipine were obtained from Sigma-Aldrich (St. Louis, USA). Tetrodotoxin (TTX) was obtained from Aik Moh (Singapore). The stock solutions of ACO (25 mM), Bay K-8644 (1 mM) and nifedipine (1 mM) were prepared in ethanol and stored at 4°C. The stock solution of TTX (5 mM) was prepared in extracellular solution and stored at -80°C.

Conventional patch-clamp recordings of cardiomyocytes

Cardiomyocytes used in this study were hiPSC-CMs purchased from Cellular Dynamics International Inc (Madison, USA). Cells were maintained at 37°C in a humidified CO2 (5%) incubator in maintenance medium provided by the company.

Whole cell configuration of the patch-clamp technique was used to measure action potentials (AP) and calcium and sodium currents. The signal was amplified using an Axon 700B patch-clamp amplifier (Axon Instrument, Sunnyvale, USA) and low-pass filtered at 5 kHz. Patch pipettes were fabricated from glass capillaries (O.D, 1.5mm; I.D, 0.9mm) using a Sutter P-97 microelectrode puller (Novato, CA, USA) and the tips were heat polished with a microforge (NARISHIGE MF-900) to gain a resistance of 2–4 MΩ. The electrical signals were sampled at 2.5–10 kHz and filtered at 2 kHz using a low-pass filter. Data acquisition was achieved using the Digidata 1440A (Axon Instrument). Data analysis and fit were performed using clamp fit 10.2 (Axon Instrument) and Origin 7.0 software (Origin Lab Corporation). A pClamp software (Version8.1; Axcon Instrument) was used to generate voltage-pulse protocols, acquire and analyze data.

Action potential recordings in hiPSC-CMs.

The APs were recorded under current-clamp mode in normal Tyrode’s solution contained (in mM): NaCl 140, KCl 5.4, CaCl2 1.8, MgCl2 1, glucose 10, and HEPES 10, adjusted to pH 7.4 with NaOH. Pipette solution contained (in mM): KCl 130, NaCl 5, MgCl2 1, MgATP 3, EGTA 10, and HEPES 10, adjusted to pH 7.2 with KOH. The parameters of APs include AP durations (APD) at 30%, 50% and 90% of repolarization (APD30, APD50, and APD90), AP amplitude (APA), maximal diastolic potential (MDP), maximal upstroke velocity (dV/dtMax) and beating frequency (BF) were analyzed [18,19]. The APDs were corrected by heart rates with Fridericia’s formula (APDc = APD/interspike interval1/3) [24]. Cells were maintained at 35°C by a temperature controller (Warner Instruments, Hamden, USA) during the recording of action potential.

L-type calcium current (ICa,L) recording in hiPSC-CMs.

Patch pipette solution contained (in mM): CsCl 120, MgCl2 3, MgATP 5, EGTA 10, and HEPES 5, adjusted to pH 7.2 with CsOH. External solution contained (in mM): NaCl 140, CsCl 10, CaCl2 1.8, MgCl2 1, glucose 10, and HEPES 10, adjusted to pH 7.4 with NaOH. To eliminate the ‘run-down’ effect during ICa,L recordings, Ba2+ was also used in the external solution (BaCl2 1.8 mM) to replace Ca2+ as charge carrier of calcium channel current [25]. Current-Voltage curve were generated by voltage-clamp protocols consisting of Vhold = -80 mV followed by a 3s long pre-pulse at -50 mV to inactivate Na+ and T-type Ca2+ channels, then a family of 300 ms depolarization from -50 mV to 50 mV in 10 mV increments.

Calcium channel current densities were obtained by dividing current amplitudes by membrane capacitances. Steady state (SS) inactivation variables of ICa,L were determined using a two-pulse gapped protocol. Potential was held at -40 mV, then pulsed to a conditioning pre-pulse ranging from -80mV to +10mV for 2000 ms, returned to -40 mV for 10 ms, and stepped to 0 mV for 250 ms at 10 s intervals. Voltage-dependence of activation curve and SS-inactivation curve were fitted with Boltzman equation (G = Gmax×[1+exp(V1/2-V)/κ]-1), where G is the conductance at various test potentials and was calculated from the peak current according to G = I/(V-Vrev), Vrev is the reversal potential obtained by extrapolating the linear part of the I/V curve to its intersection with the voltage axis. Gmax is maximum conductance; V1/2 and κ are half-activation voltage and the slope factor. The concentration-response data were fitted with Hill equation: I/Imax = 1/[1+(D/IC50)n], where I is the peak current in various concentrations of compound, Imax is the maximal peak current, D is the compound concentration, IC50 is the drug concentration for 50% inhibition, and n is the Hill coefficient.

The time course of recovery from inactivation of ICa,L was studied using a two-pulse protocol: a 250-ms pre-pulse (P1) at 0 mV from the holding potential of -50 mV followed by a variable recovery period and a 250-ms test pulse (P2) at 0 mV to assess the amount of current recovered. Each two-pulse sequence was separated by a 30 s interval. The time course of recovery for ICa,L was determined by fitting the data points to a single exponential function: I/Imax = 1-exp(-t/τ), where Imax and I were the peak current at pre-pulse (P1) and test pulse (P2), respectively; t was the variable recovery time; τ was the recovery time constant.

Peak and late sodium current (INa and INaL) recording in hiPSC-CMs and Guinea-pig ventricular myocytes.

See S1 File (Supplemental Methods).

INa, IKr and ICa,L recording in the heterologous expression systems by automated patch-clamping technique

See S1 File (Supplemental Methods) [26].

Laser-scanning confocal calcium imaging

[Ca2+]i transients were recorded in hiPSC-CMs using a LSM-710 laser-scanning confocal microscope (Carl Zeiss, Inc, Germany) with a 40×, 1.3 numerical aperture oil immersion objective and axial resolutions of 1.5 μm. Briefly, hiPSC-CMs were loaded with 2 μM Fluo-8 AM (AAT Bioquest, Inc. Sunnyvale, CA, USA) for 15 min at 37°C, and recorded in normal Tyrode’s solution. Fluo-8 was excited at 488 nm, and fluorescence emission was measured at 505 nm. Images were acquired in the line-scan (X-T) mode with 512 pixels (pixel intervals of 0.15 μm) per line at a rate of 3 ms per scan. The [Ca2+]i transients were analyzed using a modified version of MATLAB program. The Ca2+ fluorescence emission intensity was expressed as F/F0 where the F0 was the basal fluorescence intensity level. The recording was performed at 35°C temperature.

The microelectrode arrays (MEA) assay on hiPSC-CMs

See S1 File (Supplemental Methods) [24].

Statistical analysis

Values presented are means ± standard error of means (SEM) for electrophysiology assays and means ± standard derivation (SD) for calcium imaging assay. Statistical comparisons were made using Student’s unpaired or paired t-test. A value of p<0.05 was considered statistically significant.

Results

The current study adopted concentrations of ACO ranged from 0.03 to 3 μM which is consistent with that (0.1 ~10 μM) adopted in the previous in vitro studies [7~10] and it covered the blood concentrations (0.02 ~ 0.11 μM) identified in patients with ACO-induced lethal cardiac arrhythmia [27].

Effects of ACO on APs of hiPSC-CMs

Effects of ACO on APs were measured in ventricular (V)-like hiPSC-CMs which were characterized based on their AP properties including APD, APA and dV/dtMax [18,19,28]. ACO (0.3, 1.0 and 3.0 μM) concentration-dependently increased the BF, shortened APD30, APD50 and ADP90, and reduced APA in hiPSC-CMs while prolonged exposure (180~300 seconds) to a higher concentration of ACO (3.0 μM) led to more remarkable effects including a positively shift of the MDP, increased prevalence of DADs followed by VF-like AP waveforms till their complete diminishment (Fig 1A, Table 1, data for ACO at 1.0 μM not shown). The effects of ACO resembled that of nifedipine, a specific LTCC blocker. Nifedipine (0.3 μM) increased beating frequency, shortened APDs and triggered DADs in hiPSC-CMs followed by a complete abolishing of APs (Fig 1B).

[Figure omitted. See PDF.]

Fig 1. Effects of ACO on action potentials of hiPSC-CMs.

(A) Representative AP waveforms recorded in a hiPSC-CM with 120 seconds (s) exposure of ACO at 0.3 and 3.0 μM and 180s and 300s exposure of ACO at 3.0 μM. Increased beating frequencies and decreased APD and APA were observed with short-term (120s) ACO treatment. Prolonged treatment of ACO at 3.0 μM led to frequent DADs and rapid irregular VF-like AP changes. (B),Representative AP waveforms recorded in a hiPSC-CM exposed to 0.3 μM nifedipine for 60s ~ 240s. Increased beating frequencies and decreased APD occurred at 60s while DADs emerged at 180s. APs were suppressed at 240s. The arrows indicate DADs.

https://doi.org/10.1371/journal.pone.0168435.g001

[Figure omitted. See PDF.]

Table 1. Effects of ACO on action potentials of hiPSC-CMs.

https://doi.org/10.1371/journal.pone.0168435.t001

Our data demonstrated nifedipine-like effects of ACO in hiPSC-CMs characterized by shortening of APDs followed by DAD and VF-like changes.

Effects of ACO on [Ca2+]i transients in hiPSC-CMs

In hiPSC-CMs, [Ca2+]i transients reflect calcium-induced calcium release (CICR) triggered by ICa,L influx [29] and closely correlate with action potentials [30,31]. To validate the nifedipine-like effects of ACO on APs, [Ca2+]i transients were recorded in hiPSC-CMs. ACO at 0.3, 1.0 and 3.0 μM increased the frequency and decreased the duration and amplitude of [Ca2+]i transients and such effects were partially abolished by Bay K-8644, a LTCC opener. Fig 2 demonstrates the effects of ACO at 1.0 μM. Long time (180~300 seconds) exposure of ACO at 3.0 μM completely abolished the [Ca2+]i transients (data not shown).

[Figure omitted. See PDF.]

Fig 2. Effects of ACO on [Ca2+]i transients in hiPSC-CMs.

(A) Representative traces of the [Ca2+]i transients recorded in a hiPSC-CM at baseline and with subsequent exposure of ACO and ACO plus Bay K-8644. (B, C, and D) Bar-graphs show the changes in [Ca2+]i transient frequency, duration and amplitude in hiPSC-CMs. †p < 0.01, vs. baseline. ‡ p < 0.001, vs. ACO. n = 7.

https://doi.org/10.1371/journal.pone.0168435.g002

Data from [Ca2+]i transient assay further consolidated that the effects of ACO on APs are most likely to be achieved through inhibition of LTCC and CICR.

Effects of ACO on INa and ICa,L in hiPSC-CMs

Our data of AP and [Ca2+]i transients indicated that LTCC inhibition by ACO could be a major ionic change in human cardiomyocytes, which is different from the cardiac Na+ channel activation observed in animal cardiac myocytes. To validate the changes in cardiac ion currents by ACO, INa and ICa,L were measured in ACO-treated hiPSC-CMs.

Effects of ACO on INa in hiPSC-CMs.

The peak INa and late INa (INaL) were measured. ACO of up to10.0 μM exerted no significant effects on the density of INa (Fig 3Aa and 3Ab). Neither did ACO alter the SS-activation and -inactivation of INa (Fig 3Ac, Table 2) which also reflect the window currents. Moreover, ACO failed to alter INaL (Fig 3B).

[Figure omitted. See PDF.]

Fig 3. Effects of ACO on sodium currents in hiPSC-CMs.

(Aa) A representative time course and corresponding traces of the peak INa recorded in a hiPSC-CM showing the effect of ACO and TTX. (Ab) The I-V curve showing peak INa density in hiPSC-CMs. n = 7. (Ac) The SS-inactivation and -activation curves. n = 7. (Ba) A representative time course and the corresponding traces of late INa (INaL) recorded in a hiPSC-CM showing the effects of ACO and TTX. (Bb) A bar-graph showing the average densities of the INaL (at 200ms) recorded in hiPSC-CMs treated with 3 μM ACO. n = 7.

https://doi.org/10.1371/journal.pone.0168435.g003

[Figure omitted. See PDF.]

Table 2. Effect of ACO on SS-activation and -inactivation of INa in hiPSC-CMs.

https://doi.org/10.1371/journal.pone.0168435.t002

Effects of ACO on ICa,L in hiPSC-CMs.

With the “run-down” phenomenon minimized after replacement of extracellular Ca2+ with Ba2+, the inhibitory effect of ACO on ICa,L was confirmed and validated by addition of nifedipine which further blocked ICa,L (Fig 4A).The inhibitory effect of ACO on LTCC was further validated by addition of Bay K-8644 (1.0 μM), an ICa,L opener, which reversed the effect of ACO (Fig 4B). Moreover, a positive shift of the half-activation voltage (V1/2) and a negative shift of the V1/2 of ICa,L were observed in hiPSC-CMs treated with ACO while the recovery from inactivation remained unchanged (Fig 4C, 4D and 4E, Table 3). hiPSC-CMs exposed to 0.1 ~ 10.0 μM of ACO demonstrated a strong concentration-dependent inhibitory effect on ICa,L with an IC50 of 0.68 ± 0.15 μM (Fig 4F).

[Figure omitted. See PDF.]

Fig 4. Effects of ACO on ICa,L in hiPSC-CMs.

(A) Representative traces of the voltage-gated LTCC current ICa,L recorded from a hiPSC-CM at baseline followed by exposure of 1 μM ACO and addition of 1 μM nifedipine. (B) Representative traces and I-V curves recorded from hiPSC-CMs exposed to 1 μM ACO followed by 1 μM Bay K-8644. *p < 0.05, †p < 0.01, ‡ p < 0.001, ACO vs. baseline or Bay K-8644 vs ACO, n = 8. (C) The voltage-dependent activation curve of ICa,L recorded in hiPSC-CMs exposed to 1 μM of ACO. G/Gmax: normalized peak conductance. *p < 0.05, †p < 0.01, vs. baseline. n = 5. (D) The SS-inactivation curve of ICa,L. *p < 0.05, †p < 0.01, ‡ p < 0.001, vs. baseline. n = 5. (E) Recovery from inactivation of ICa,L. n = 5. Insert shows the 2-pulse protocol. (F) The IC50 of ACO on ICa,L (n = 8).

https://doi.org/10.1371/journal.pone.0168435.g004

[Figure omitted. See PDF.]

Table 3. Effect of ACO on the kinetics of ICa,L

https://doi.org/10.1371/journal.pone.0168435.t003

Effects of ACO on INa, IKr and ICa,L recorded in the heterologous expression system

Effects of ACO on major ion currents in hiPSC-CMs were validated in the heterologous expression systems. Using the Patchliner® automated patch-clamping system, effects of ACO (0.01 ~100 μM) on INa, ICa,L IKr were measured in Nav1.5-HEK293, Cav1.2-CHO and hERG-HEK293 cells, respectively. ACO showed milder blockage of INa and IKr in Nav1.5-HEK293 and hERG-HEK293 cells with IC50s of 14.1 μM (n = 14) and 16.0 μM (n = 19), respectively (Fig 5A and 5B). However, a stronger ICa,L blocking was noted in Cav1.2-CHO cells exposed to lower concentrations of ACO with IC50 of 6.6 μM (n = 5) (Fig 5C).

[Figure omitted. See PDF.]

Fig 5. Effects of ACO on INa, IKr and ICa,L in heterologous expression systems.

Nav1.5-HEK293, hERG-HEK293 and Cav1.2-CHO cells were exposed to 0.1~100 μM ACO. (A) The concentration-dependent inhibition curve of INa and representative traces recorded. (B) The concentration-dependent inhibition curve of IKr (peak tail: the peak of the tail current) and representative traces. (C) The concentration-dependent inhibition curve of ICa,L and representative traces.

https://doi.org/10.1371/journal.pone.0168435.g005

Effects of ACO on extracellular filed potentials

To validate the potential impact of ACO-induced APD shortening on electrocardiography (ECG), the extracellular field potentials (FP) of hiPSC-CMs were measured using the MEA system which produces signals resemble the ECG. ACO-treated hiPSC-CMs showed a tendency of concentration-dependent increase of BF, decrease of Ca2+ wave and shortening of FPDc (Fig 6). Yet the heart rate increments and Ca2+ wave reduction were obvious at higher concentrations of ACO while a moderate FPDc shortening was significant at 0.03 and 0.1 μM of ACO (Fig 6B). Cell contractions were stopped by ACO at higher concentrations and with longer exposure time (> 3 μM and > 300s) (data not shown).

[Figure omitted. See PDF.]

Fig 6. Effects of ACO on the field potentials in hiPSC-CMs.

(A) Representative traces of the filed potentials (FD) recorded with a cluster of hiPSC-CMs at baseline, exposed to various concentrations of ACO and after washout. (B) A bar-graph demonstrates the average beating frequency (BF), Ca2+ waves and FPDc recorded at baseline, with 0.03, 0.1, 0.3, 1.0 and 3.0 μM of ACO treatments and after washout. *p< 0.05, ‡p< 0.001; vs. baseline. n = 4.

https://doi.org/10.1371/journal.pone.0168435.g006

Data from the MEA assay, including the increased BF, decreased Ca2+ wave (which echoes ICa,L inhibition observed with single cells) and shortened FPDc, support our findings in single hiPSC-CM.

Effects of ACO on INa in isolated Guinea-pig ventricular myocytes

To validate the effect of ACO on sodium currents in hiPSC-CMs, INa and INaL were measured in isolated Guinea-pig ventricular myocytes treated with 3 μM ACO. The peak sodium current densities recorded during -80mV to -50mV significantly increased by ACO while it remained unchanged during -45 mV to +20 mV (Fig 7A and 7B). Moreover, ACO negatively shifted the activation of INa by 12.58 mV (-47.66 ± 0.71 vs -60.24 ± 2.22) and positively shifted the inactivation of INa by 4.32 mV (-69.97 ± 1.78 vs -74.29 ± 1.20) (Fig 7C, Table 4) and markedly increased the INaL at the voltages between -55 and -85 mV (Fig 7D and 7E). Thus, the well-anticipated effect of ACO on opening of the INaL was validated in Guinea pig cardiac myocytes.

[Figure omitted. See PDF.]

Fig 7. Effects of ACO on sodium currents in Guinea-pig ventricular myocytes.

(A) Representative traces of the peak INa recorded in a Guinea-pig ventricular myocyte prior to and after 3 μM ACO. (B) The I-V curves showing the current densities of peak INa. n = 5. (C) The SS-inactivation and -activation curves. n = 5. (D) The I-V curves of the voltage-gated INaL densities. n = 5. (E) Representative traces of voltage-dependent INaL recorded in a Guinea-pig ventricular myocyte prior to and post 3μM ACO. *p < 0.05, †p < 0.01, vs. baseline.

https://doi.org/10.1371/journal.pone.0168435.g007

[Figure omitted. See PDF.]

Table 4. Effect of ACO on SS-activation and -inactivation of INa in Guinea-pig ventricular myocytes.

https://doi.org/10.1371/journal.pone.0168435.t004

Discussion

In the current study, we observed that ACO, within the well-adopted in vitro doses and closer to the toxicological concentrations, is capable of triggering proarrhythmic effects in human cardiomyocytes characterized by increased beating frequency and shortened repolarization phases of action potentials and field potentials followed by increased prevalence of DADs and cardiac arrest. Such changes were accompanied by reduced duration and amplitude of [Ca2+]i transients and Ca2+ wave. Moreover, data from hiPSC-CMs and the heterologous expression systems directly indicated that ACO could selectively inhibit ICa,L independent from [Na+]i and Na+ channel activities as the peak, late and window INa remained unaffected.

ACO has long been known as a neurotoxin that selectively interacts with the voltage-dependent neuronal Na+ channels by binding to its neurotoxin binding site 2 at the pore region of the alpha subunit, rendering a prolonged opening of the Na+ channel [32]. In isolated cardiac myocytes of various animals, ACO has been shown to act as an INa agonist that opens the Na+ channels during the depolarization/repolarization phases, leading to a delayed repolarization [7~9]. However, our data from human hiPSC-CMs and transgenic cell line (Nav1.5-HEK293), controlled by Guinea-pig ventricular myocytes, minimized the possibility that ACO-increased cytosolic [Na+]i in human cardiomyocytes acts as the main driver of cardiac arrhythmia [13,14]. Rather, our data indicate that inhibitions of LTCC and CICR by ACO could play a key role in the proarrhythmic effects of ACO in human and ACO is capable of blocking LTCC following a mechanism independent from increased [Na+]i. The effects of ACO on cardiac repolarization in hiPSC-CMs resemble that of nifedipine, which, like Bay K-8644, belongs to dihydropyridines, the most selective and potent calcium channel modulators specifically identify LTCC activity in cardiac myocyte [33]. Moreover, the effect of ACO on beating frequency also resemble that of nifedipine that has been shown capable of increasing the BF in cardiomyocytes derived from hiPSCs and human embryonic stem cells [34,35].

Defects in cardiac LTCC have been associated with severe arrhythmias in human. Loss-of-function of cardiac LTCC due to genetic mutations have been associated with inherited J-wave syndromes and sudden cardiac death [36] and short QT syndrome characterized by an abnormally short QT interval and an increased risk of atrial and ventricular fibrillation [25,37]. Decreased ICa,L has been found in patients with atrial fibrillation [38].

Nevertheless, caution should be exercised to apply the data obtained in hiPSC-CMs to the intact human heart. ECG findings in patients with ACO-induced cardiotoxicity have been explained so far by the persistent activation of Na+ channels and subsequent accumulation of cytosolic Ca2+. The current study, however, suggests that ICa,L inhibition plays a key role in the effects of ACO in human cardiomyocytes. We noted that ACO at lower concentrations and shorter exposure time (< 3 μM and < 300s) could rapidly shorten the repolarization phase (APD and FPDc) in hiPSC-CMs while increased dosages and longer exposure time abolished the electrical activities (Figs 2A and 6). In patients with aconitine intoxication, ICa,L inhibition, rather than INa activation, may offer a new explanation for atrioventricular blocks and bradycardia, the major ECG defects identified in almost all patients, in addition to the effect of parasympathetic activation, as ICa,L is the major driver of the depolarization of the pacemaker myocytes. However, data from hiPSC-CMs fall short to explain why QT-shortening has not been found in patients yet QT prolongation has been reported in some patients instead [39]. While the neurohormonal effects of ACO in cardiotoxicity [1] may explain some of the discrepancy between the electrophysiological characteristics of a ventricular myocyte (such as APs) and the heart (such as ECG), a possible role of IKr inhibition, the doses of ACO and the exposure time/duration might all contribute to the variable cardiac repolarization durations in human heart. Firstly, we recorded a moderate IKr inhibition in hERG-HEK293 cells treated by ACO which is consistent with previous in vitro studies [16,17]. Under certain conditions, the effects of IKr inhibition may become dominant over that of ICa,L. Secondly, different species of ACO-containing herbs and different ways of administration could contribute to different concentrations of ACO in patients. Lastly, our data from hiPSC-CMs were obtained within 30 minutes while it takes an average of 4 hours (earliest 2 hours) to have the 1st ECG recorded in patients with ACO ingestions [1], leaving the early change of ECG unchecked.

In addition, our data hint that ACO may also inhibit LTCC in vascular smooth muscle cells and compromise the vascular tone and thus contribute to the severe hypotensive effect in addition to its actions on autonomic nervous system. It is reported that overdose of LTCC inhibiters could cause severe hypotension [40] likely due to reduced vascular tones following the LTCC inhibition in vascular smooth muscle cells [41].

Although blocking of Kv1.5 channels and IKur in ACO-treated neonatal cardiomyocytes have been associated with prolonged APDs and EADs [15,16], such changes are unlikely to happen in human ventricular myocytes and pacemaker cells as Ikur is predominantly expressed in atrial myocytes [42].

Aconitum, along or as an important ingredient in many traditional Chinese medicine prescriptions, will continually be adopted for treating various disease conditions and ACO-induced arrhythmias will remain to be a health threat [1,2]. To date, there is no effective ion channel-targeted therapy for ACO poisoning as sodium channel blockers appeared to be the least effective drugs [1]. Our data may shed new light on more targeted drug therapies.

The current study highlights the importance of applying human cardiomyocyte models for evaluating the effects of various drugs. The identification of a different mechanism of ACO-induced arrhythmia in human cardiomyocytes compared to that from other animals, particularly rodents, highlights the intra-species variations in cardiac electrophysiology [22, 23]. Data of the current and previous studies indicate that the cardiac Na+ channels of different species could have different sensitive to ACO while the TTX-insensitive Na+ channels could be insensitive to ACO as a high concentration of ACO (100 μM) [43] on human Na+ channels achieved the similar effect of 3 μM ACO (close to physiological concentrations) on isolated Guinea-pig ventricular myocytes. Whereas lower concentrations of ACO of (3~10 μM) failed to alter human INa.

In summary, for the first time we observed the proarrhythmic effects of ACO in human cardiomyocytes and such effects are likely to be achieved via inhibition of LTCC.

Supporting Information

[Figure omitted. See PDF.]

S1 File. Supplemental Methods.

https://doi.org/10.1371/journal.pone.0168435.s001

(DOC)

Author Contributions

1. Conceptualization: HW SS.

2. Data curation: HW SS.

3. Formal analysis: JW XW YYC CHK ZL HW SS.

4. Funding acquisition: HW SS CW.

5. Investigation: JW XW YYC CHK ZL HG.

6. Methodology: HW SS JW XW YYC CHK HG.

7. Resources: HW SS CW.

8. Software: QY.

9. Supervision: HW SS CW.

10. Validation: ZL.

11. Writing – original draft: HW.

12. Writing – review & editing: HW JW XW SS.

Citation: Wu J, Wang X, Chung YY, Koh CH, Liu Z, Guo H, et al. (2017) L-Type Calcium Channel Inhibition Contributes to the Proarrhythmic Effects of Aconitine in Human Cardiomyocytes. PLoS ONE 12(1): e0168435. https://doi.org/10.1371/journal.pone.0168435

References

1. Tai YT, But PP, Young K, Lau CP (1992) Cardiotoxicity after accidental herb-induced aconite poisoning. Lancet 340: 1254–1256. pmid:1359321

2. Chan TY (2009) Aconite poisoning. Clin Toxicol (Phila) 47: 279–285.

3. Colussi C, Berni R, Rosati J, Straino S, Vitale S, Spallotta F, et al. (2010) The histone deacetylase inhibitor suberoylanilide hydroxamic acid reduces cardiac arrhythmias in dystrophic mice. Cardiovasc Res 87: 73–82. pmid:20164117

4. Bartosova L, Novak F, Bebarova M, Frydrych M, Brunclik V, Opatrilova R, et al. (2007) Antiarrhythmic effect of newly synthesized compound 44Bu on model of aconitine-induced arrhythmia—compared to lidocaine. Eur J Pharmacol 575: 127–133. pmid:17706639

5. Amran MS, Hashimoto K, Homma N (2004) Effects of sodium-calcium exchange inhibitors, KB-R7943 and SEA0400, on aconitine-induced arrhythmias in guinea pigs in vivo, in vitro, and in computer simulation studies. J Pharmacol Exp Ther 310: 83–89. pmid:15028781

6. Ohno Y, Chiba S, Uchigasaki S, Uchima E, Nagamori H, Mizugaki M, et al. (1992) The influence of tetrodotoxin on the toxic effects of aconitine in vivo. Tohoku J Exp Med 167: 155–158. pmid:1475787

7. Peper K, Trautwein W (1967) The effect of aconitine on the membrane current in cardiac muscle. Pflugers Archiv fur die gesamte Physiologie des Menschen und der Tiere 296: 328–36.

8. Nilius B, Boldt W, Benndorf K (1986) Properties of aconitine-modified sodium channels in single cells of mouse ventricular myocardium. Gen Physiol Biophys 5: 473–484. pmid:2433183

9. Honerjager P, Meissner A (1983) The positive inotropic effect of aconitine. Naunyn Schmiedebergs Arch Pharmacol 322: 49–58. pmid:6302523

10. Liu QH, Li XL, Xu YW, Lin YY, Cao JM, Wu BW (2012) A novel discovery of IK1 channel agonist: zacopride selectively enhances IK1 current and suppresses triggered arrhythmias in the rat. J Cardiovasc Pharmacol 59: 37–48 pmid:21921806

11. Fu M, Li RX, Fan L, He GW, Thornburg KL, Wang Z. (2008) Sarcoplasmic reticulum Ca2+ release channel ryanodine receptor (RyR2) plays a crucial role in aconitine-induced arrhythmias. Biochem Pharmacol 75: 2147–2156. pmid:18439986

12. Chen L, Ma C, Cai BC, Lu YM, Wu H (1995) [Single channel analysis of aconitine blockade of calcium channels in rat myocardiocytes]. Yao Xue Xue Bao 30: 168–171. pmid:7543723

13. Findlay I (2004) Physiological modulation of inactivation in L-type Ca2+ channels: one switch. J Physiol 554: 275–283 pmid:12824441

14. Hancox JC, Levi AJ (1996) Actions of the digitalis analogue strophanthidin on action potentials and L-type calcium current in single cells isolated from the rabbit atrioventricular node. Br J Pharmacol 118:1447–1454. pmid:8832071

15. Wang YJ, Chen BS, Lin MW, Lin AA, Peng H, Sung RJ, et al. (2008) Time-dependent block of ultrarapid-delayed rectifier K+ currents by aconitine, a potent cardiotoxin, in heart-derived H9c2 myoblasts and in neonatal rat ventricular myocytes. Toxicol Sci 106: 454–463. pmid:18779382

16. Li Y, Tu D, Xiao H, Du Y, Zou A, Liao Y, et al. (2010) Aconitine blocks HERG and Kv1.5 potassium channels. J Ethnopharmacol 131: 187–195. pmid:20600762

17. Xie S, Jia Y, Liu A, Dai R, Huang L (2015) Hypaconitine-induced QT prolongation mediated through inhibition of KCNH2 (hERG) potassium channels in conscious dogs. J Ethnopharmacol 166: 375–379. pmid:25800797

18. Moretti A, Bellin M, Welling A, Jung CB, Lam JT, Bott-Flügel L, et al. (2010) Patient-specific induced pluripotent stem-cell models for long-QT syndrome. N Engl J Med 363: 1397–1409. pmid:20660394

19. Ma D, Wei H, Zhao Y, Lu J, Li G, Sahib NB, et al. (2013) Modeling type 3 long QT syndrome with cardiomyocytes derived from patient-specific induced pluripotent stem cells. Int J Cardiol 168: 5277–5286. pmid:23998552

20. Liang P, Lan F, Lee AS, Gong T, Sanchez-Freire V, Wang Y, et al. (2013) Drug screening using a library of human induced pluripotent stem cell-derived cardiomyocytes reveals disease-specific patterns of cardiotoxicity. Circulation 127: 1677–1691. pmid:23519760

21. Lu J, Wei H, Wu J, Jamil MF, Tan ML, Adenan MI, et al. (2014) Evaluation of the cardiotoxicity of mitragynine and its analogues using human induced pluripotent stem cell-derived cardiomyocytes. PLoS One 9: e115648 pmid:25535742

22. Linz KW, Meyer R (2000) Profile and kinetics of l-type calcium current during the cardiac ventricular action potential compared in guinea-pigs, rats and rabbits. Pflugers 439: 588–599.

23. O'Hara T, Rudy Y (2012) Quantitative comparison of cardiac ventricular myocyte electrophysiology and response to drugs in human and nonhuman species. Am J Physiol Heart Circ Physiol 302: H1023–1030. pmid:22159993

24. Clements M, Thomas N (2014) High-throughput multi-parameter profiling of electrophysiological drug effects in human embryonic stem cell derived cardiomyocytes using multi-electrode arrays. Toxicol Sci 140: 445–461. pmid:24812011

25. Templin C, Ghadri JR, Rougier JS, Baumer A, Kaplan V, Albesa M, et al. (2011) Identification of a novel loss-of-function calcium channel gene mutation in short QT syndrome (SQTS6). Eur Heart J 32: 1077–1088. pmid:21383000

26. Rajamohan D, Kalra S, Duc Hoang M, George V, Staniforth A, Russell H, et al. (2016) Automated Electrophysiological and Pharmacological Evaluation of Human Pluripotent Stem Cell-Derived Cardiomyocytes. Stem Cells Dev 25: 439–452. pmid:26906236

27. Niitsu H, Fujita Y, Fujita S, Kumagai R, Takamiya M, Aoki Y, et al. (2013) Distribution of Aconitum alkaloids in autopsy cases of aconite poisoning. Forensic Sci Int 227: 111–117. pmid:23131306

28. Ma J, Guo L, Fiene SJ, Anson BD, Thomson JA, Kamp TJ, et al. (2011) High purity human-induced pluripotent stem cell-derived cardiomyocytes: electrophysiological properties of action potentials and ionic currents. Am J Physiol Heart Circ Physiol 301: H2006–2017. pmid:21890694

29. Zhang GQ, Wei H, Lu J, Wong P, Shim W (2013) Identification and characterization of calcium sparks in cardiomyocytes derived from human induced pluripotent stem cells. PLoS One 8: e55266. pmid:23408964

30. Spencer CI, Baba S, Nakamura K, Hua EA, Sears MA, Fu CC, et al. (2014) Calcium transients closely reflect prolonged action potentials in iPSC models of inherited cardiac arrhythmia. Stem Cell Reports 3: 269–281. pmid:25254341

31. Kim JJ, Yang L, Lin B, Zhu X, Sun B, Kaplan AD, et al. (2015) Mechanism of automaticity in cardiomyocytes derived from human induced pluripotent stem cells. J Mol Cell Cardiol 81: 81–93. pmid:25644533

32. Rao S, Sikdar SK (2000) Modification of alpha subunit of RIIA sodium channels by aconitine. Pflugers Arch 439: 349–355. pmid:10650987

33. Bers DM, Perez-Reyes E (1999) Ca channels in cardiac myocytes: structure and function in Ca influx and intracellular Ca release. Cardiovasc Res 42: 339–360. pmid:10533572

34. Sheng X, Reppel M, Nguemo F, Mohammad FI, Kuzmenkin A, Hescheler J, et al. (2012) Human pluripotent stem cell-derived cardiomyocytes: response to TTX and lidocain reveals strong cell to cell variability. PLoS One 7: e45963. pmid:23029342

35. Harris K, Aylott M, Cui Y, Louttit JB, McMahon NC, Sridhar A (2013) Comparison of electrophysiological data from human-induced pluripotent stem cell-derived cardiomyocytes to functional preclinical safety assays. Toxicol Sci 134: 412–426. pmid:23690542

36. Burashnikov E, Pfeiffer R, Barajas-Martinez H, Delpon E, Hu D, Desai M, et al. (2010) Mutations in the cardiac L-type calcium channel associated with inherited J-wave syndromes and sudden cardiac death. Heart Rhythm 7: 1872–1882. pmid:20817017

37. Antzelevitch C, Pollevick GD, Cordeiro JM, Casis O, Sanguinetti MC, Aizawa Y, et al. (2007) Loss-of-function mutations in the cardiac calcium channel underlie a new clinical entity characterized by ST-segment elevation, short QT intervals, and sudden cardiac death. Circulation 115: 442–449. pmid:17224476

38. Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM (1999) Atrial L-type Ca2+ currents and human atrial fibrillation. Circ Res 85: 428–436. pmid:10473672

39. Kim TK, Jin SC, Kim SJ, Choi WI (2013) Analysis of QTc Prolongation Related to Arrhythmia in Patients with Aconitine Intoxication. Journal of The Korean Society of Emergency Medicine 24: 548–556.

40. Belson MG, Gorman SE, Sullivan K, Geller RJ (2000) Calcium channel blocker ingestions in children. Am J Emerg Med 18: 581–586. pmid:10999574

41. Zhao Q, Zhao KS (2007) Inhibition of L-type calcium channels in arteriolar smooth muscle cells is involved in the pathogenesis of vascular hyporeactivity in severe shock. Shock 28: 717–721. pmid:17607159

42. Ravens U, Wettwer E (2011) Ultra-rapid delayed rectifier channels: molecular basis and therapeutic implications. Cardiovasc Res 89: 776–785. pmid:21159668

43. Wright SN (2002) Comparison of aconitine-modified human heart (hH1) and rat skeletal (mu1) muscle Na+ channels: an important role for external Na+ ions. J Physiol 538: 759–71. pmid:11826163