Assessment of the proarrhythmic effects of repurposed antimalarials for COVID-19 treatment using a comprehensive in vitro proarrhythmia assay (CiPA)

Due to the outbreak of the SARS-CoV-2 virus, drug repurposing and Emergency Use Authorization have been proposed to treat the coronavirus disease 2019 (COVID-19) during the pandemic. While the efficiency of the drugs has been discussed, it was identified that certain compounds, such as chloroquine and hydroxychloroquine, cause QT interval prolongation and potential cardiotoxic effects. Drug-induced cardiotoxicity and QT prolongation may lead to life-threatening arrhythmias such as torsades de pointes (TdP), a potentially fatal arrhythmic symptom. Here, we evaluated the risk of repurposed pyronaridine or artesunate-mediated cardiac arrhythmias alone and in combination for COVID-19 treatment through in vitro and in silico investigations using the Comprehensive in vitro Proarrhythmia Assay (CiPA) initiative. The potential effects of each drug or in combinations on cardiac action potential (AP) and ion channels were explored using human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) and Chinese hamster ovary (CHO) cells transiently expressing cardiac ion channels (Nav1.5, Cav1.2, and hERG). We also performed in silico computer simulation using the optimized O’Hara-Rudy human ventricular myocyte model (ORd model) to classify TdP risk. Artesunate and dihydroartemisinin (DHA), the active metabolite of artesunate, are classified as a low risk of inducing TdP based on the torsade metric score (TMS). Moreover, artesunate does not significantly affect the cardiac APs of hiPSC-CMs even at concentrations up to 100 times the maximum serum concentration (Cmax). DHA modestly prolonged at APD90 (10.16%) at 100 times the Cmax. When considering Cmax, pyronaridine, and the combination of both drugs (pyronaridine and artesunate) are classified as having an intermediate risk of inducing TdP. However, when considering the unbound concentration (the free fraction not bound to carrier proteins or other tissues inducing pharmacological activity), both drugs are classified as having a low risk of inducing TdP. In summary, pyronaridine, artesunate, and a combination of both drugs have been confirmed to pose a low proarrhythmogenic risk at therapeutic and supratherapeutic (up to 4 times) free Cmax. Additionally, the CiPA initiative may be suitable for regulatory use and provide novel insights for evaluating drug-induced cardiotoxicity.

Due to the outbreak of the SARS-CoV-2 virus, drug repurposing and Emergency Use Authorization have been proposed to treat the coronavirus disease 2019 (COVID-19) during the pandemic.While the efficiency of the drugs has been discussed, it was identified that certain compounds, such as chloroquine and hydroxychloroquine, cause QT interval prolongation and potential cardiotoxic effects.Drug-induced cardiotoxicity and QT prolongation may lead to lifethreatening arrhythmias such as torsades de pointes (TdP), a potentially fatal arrhythmic symptom.Here, we evaluated the risk of repurposed pyronaridine or artesunate-mediated cardiac arrhythmias alone and in combination for COVID-19 treatment through in vitro and in silico investigations using the Comprehensive in vitro Proarrhythmia Assay (CiPA) initiative.The potential effects of each drug or in combinations on cardiac action potential (AP) and ion channels were explored using human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) and Chinese hamster ovary (CHO) cells transiently expressing cardiac ion channels (Nav1.5,Cav1.2, and hERG).We also performed in silico computer simulation using the optimized O'Hara-Rudy human ventricular myocyte model (ORd model) to classify TdP risk.Artesunate and dihydroartemisinin (DHA), the active metabolite of artesunate, are classified as a low risk of inducing TdP based on the torsade metric score (TMS).Moreover, artesunate does not significantly affect the cardiac APs of hiPSC-CMs even at concentrations up to 100 times the maximum serum concentration (C max ).DHA modestly prolonged at APD 90 (10.16%) at 100 times the C max .When considering C max , pyronaridine, and the combination of both drugs (pyronaridine and artesunate) are classified as having an intermediate risk of inducing TdP.However, when considering the unbound concentration (the free fraction not bound to carrier proteins or other tissues inducing pharmacological activity), both drugs are classified as having a low risk of inducing TdP.In summary, pyronaridine, artesunate, and a combination of both drugs have been confirmed to pose a low proarrhythmogenic risk at therapeutic and supratherapeutic (up to 4 times) free C max .Additionally, the CiPA initiative may be suitable for regulatory use and provide novel insights for evaluating drug-induced cardiotoxicity.

Introduction
Drug-induced QT/QTc interval prolongation and cardiotoxicity are associated with arrhythmic events, such as torsade de point (TdP), and sudden cardiac death (Yap and Camm, 2003).To evaluate these risk factors, the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) has published S7B and E14 guidelines, which provide a framework for predicting cardiotoxicity assessment for drugs (Cavero and Crumb, 2005).Unfortunately, the low specificity of ICH S7B/E14 leads to the undesirable regulation of drugs that are unlikely to cause cardiac arrhythmias (Jeong et al., 2022).This is typically understood to result from focusing on only one of the many ion channels governing action potential (AP) formation (Colatsky et al., 2016;Jeong et al., 2022).To resolve this issue, the Comprehensive in vitro Proarrhythmia Assay (CiPA) project was proposed by the US Food and Drug Administration (FDA) and has grown into an international research project with the participation of the FDA and various researchers led by pharmaceutical companies.The CiPA project comprises four interrelated components, the first of which assesses the effects of drugs through in vitro experiments using ion currents participating in human ventricular AP formation (Cavero et al., 2016).The second component predicts drug-induced cardiotoxicity using an in silico model, the O'Hara-Rudy human ventricular myocyte model (ORd model), with results obtained from an in vitro ion current dataset as input (Cavero et al., 2016;Yoo et al., 2021).The in silico model allows the classification of each drug as having no/low, intermediate, or high proarrhythmic risk.The third component investigates the electrophysiological effects of human induced pluripotent stem cells (hiPSC-CMs), which can also be used to confirm unexpected adverse cardiac effects compared with in vitro multiple cardiac ion channel screening data and in silico prediction models (Strauss et al., 2019).The fourth component confirms in vivo human electrophysiological impact of drugs that may result from metabolic characteristics through human surface electrocardiography in phase I clinical trials and compares it to preclinical data (Park et al., 2019;Strauss et al., 2019).The CiPA initiative suggests a novel human-based paradigm for systematically evaluating drugs with the potential proarrhythmic risk rather than focusing on druginduced hERG block and QT prolongation (Cavero et al., 2016;Fermini et al., 2016;Authier et al., 2017).This new paradigm can overcome the limitations of existing methods used to assess druginduced cardiotoxicity.Moreover, the newly ICH E14/S7B Q&As incorporate CiPA assays as follow-up studies for risk evaluation best practice considerations for experimental factors (ICH E14/S7B Implementation Working Group, 2022).

FIGURE 2
Effects of pyronaridine tested considering unbound concentration on cardiac ionic currents in CHO cells.(A-D) Representative traces of ionic currents demonstrating the effect of pyronaridine on Peak Nav1.5 (A), Late Nav1.5 (B), Cav1.2 (C), and hERG (D) at concentrations of 13.2, 26.4, 39.6, and 52.8 nM, respectively (left), and the concentration-response relationship for each ionic current (right).Data are presented as mean ± standard error of the mean (SEM) and analyzed using Student's t-test.*p < 0.05, **p < 0.01 or ***p < 0.001 compared to control (n = 3).et al., 2021); many nonclinical and clinical trials have been conducted to treat COVID-19, including the repurposing of drugs such as chloroquine (CQ), hydroxychloroquine (HCQ), and remdesivir (Yanagida et al., 2021).Drug repurposing is a useful tool for identifying new therapeutic methods (Serafin et al., 2020).Previous studies have reported that the quinolinerelated antimalarial drugs CQ and HCQ block ion channels related to cardiac AP formation and can cause fatal arrhythmias associated with QT prolongation (Kinoshita et al., 2010;Borba et al., 2020;Chorin et al., 2020;Mercuro et al., 2020;Wang et al., 2020;Tleyjeh et al., 2021;Sala et al., 2022).Compliant data from the CiPA initiative have shown that using CQ and HCQ at therapeutic concentrations related to malaria or repurposing for COVID-19 is associated with the risk of QT prolongation and TdP (Delaunois et al., 2021).Those drugs inhibited mainly potassium channels, in silico prediction model predicted that have the potential to prolong repolarization (Thomet et al., 2021;Varshneya et al., 2021;Whittaker et al., 2021;TeBay et al., 2022).Additionally, CQ, HCQ, or azithromycin (AZM) co-administration can trigger to high arrhythmogenic risk in conditions with impaired repolarization (Sutanto and Heijman, 2020;Montnach et al., 2021).Consequently, on 15 June 2020, the FDA revoked the Emergency Use Authorization (EUA) for HCQ and CQ because they were ineffective in treating COVID-19 and could cause serious adverse cardiac events (The Food and Drug Administration, 2020a).Pyronaridine-artesunate is also an antimalarial drug that has been considered for treating COVID-19, and clinical trials are currently ongoing.Pyronaridine-artesunate is an artemisinin-based combination therapy used to treat malaria (Rueangweerayut et al., 2012).In vitro studies comparing the effectiveness of pyronaridine, artesunate, and HCQ against SARS-CoV-2 indicated that pyronaridine and artesunate were more effective in human lung epithelial (Calu-3) cells (Bae et al., 2020;Krishna et al., 2021).However, there is limited information on drug-induced cardiotoxicity and QT prolongation associated with repurposing for COVID-19 treatment.Therefore, it is important to determine the potential cardiotoxicity of this drug to avoid severe adverse events.
Our study aimed to perform in vitro and in silico investigations of the effects of pyronaridine, artesunate, and a combination of both drugs on cardiac electrophysiological properties and to predict arrhythmic potentials using the principles and methods of the CiPA initiative.Moreover, our study aimed to improve our understanding of cardiotoxicity and QT prolongation mechanisms, providing even greater insight into standardized nonclinical assays based on the CiPA approach.

Chemicals
Pyronaridine and artesunate were obtained from Sigma-Aldrich (St. Louis, MO, USA).The mixture consists of pyronaridine and artesunate in a ratio of 3:1.Dihydroartemisinin (DHA) was obtained from Selleck Chemicals (Houston, TX, USA).The drugs were formulated into a stock solution using dimethyl sulfoxide (DMSO) and stored at −20 °C.On the day of the experiment, the stock solution was freshly diluted in the external solution containing 0.1% DMSO and the desired drug concentrations.All chemicals required for preparing the external/internal solutions were obtained from Sigma-Aldrich.The drugs were tested at four different concentrations, taking into account the maximum serum concentration after drug administered (C max ) and the concentration of the drug that is not bound to plasma or tissue proteins or not uptake by red blood cells (unbound concentration) (Table 1) (The European Medicines Agency, 2012).According to The European Medicines Agency (2012), therapeutic Cmax of pyronaridine in humans is 726 ng/mL (800 nM), and unbound concentration is 12 ng/mL (13.2 nM).Pyronaridine exhibited high plasma protein binding in various organisms, including humans, dogs, rats, and rabbits ranging from 92% to 96%, and it showed preferentially associated with red blood cells.The therapeutic Cmax of artesunate in humans is 0.3 μg/mL (780 nM).The Cmax of DHA, the active metabolite of artesunate, in humans is 1.2 μg/mL (4,200 nM).

Recording of ionic currents
Conventional whole-cell voltage-clamp recordings were performed for four different currents: hERG, peak Nav1.5, late

Recording of spontaneous APs in hiPSC-CMs
For single-cell AP recordings, hiPSC-CMs (Cardiosight-S; NEXEL, Co., Ltd., Seoul, Korea) were cultured.The cells were transferred to four-well culture plates with Matrigel (Corning, NY, USA)-coated glass coverslips for AP recording.They were maintained in a culture incubator in a humidified 5% CO 2 atmosphere at 37 °C, and utilized within 7 days after thawing.The medium was changed every 2-3 days.The external solution for APs recording contained the following (in mM): 145 NaCl, 5.4 KCl, 10 HEPES, 1 MgCl 2 *6H 2 O, 1.8 CaCl 2 *H 2 O, and 5 dextrose (pH 7.4 with NaOH).The internal solution for AP recording consisted of the following (in mM): 20 KCl, 120 K-Aspartic Acid, 5 NaCl, 2 CaCl 2 *H 2 O, 5 EGTA, and 5 Mg-ATP (pH 7.25 with KOH).The maximum rate of depolarization during the upstroke of the AP (Vmax), maximum diastolic potential (MDP), AP amplitude (APA), the AP duration at 50% (APD 50 ), and 90% (APD 90 ) repolarization were measured when they were stable.The spontaneous contractions of hiPSC-CMs were recorded in I = 0 mode, and only cells that exhibited stable beating were used in the analysis.APs were recorded at a physiological temperature (37 °C ± 1 °C).

Hill fitting and generating samples
Hill fitting was conducted using the same method as proposed by Chang et al. (2017) with R programming language (https://github.com/FDA/CiPA/tree/Model-Validation-2018/Hill_Fitting).In vitro datasets obtained from the whole-cell voltage-clamp experiment were fitted and sampled with the Markov chain Monte Carlo (MCMC) model proposed by Haario et al. (2006); Laine and Tamminen (2008); Soetaert and Petzoldt (2010) to obtain Hill curves for dosage-to-response for each ion channel.The results were 2,000 samples with half-maximal inhibitory concentration (IC50) and Hill coefficients data per drug used as input for in silico model.

In silico model
The optimized ORd model for drug effects was used as a cellular electrophysiological in silico model for the human ventricular myocytes (Dutta et al., 2017), generating the AP, the sum of the potentials for all ion channels inhibited by the drugs in cardiomyocytes (O'Hara et al., 2011).The inhibited ion channel was expressed by multiplying the inhibition factor in the general ionic current equation; the inhibition factor consists of Hill coefficients (h), IC 50 , and drug concentration (D) (equation 1) (Li et al., 2019).

Torsade metric score (TMS)
The qNet is the total amount of net charges that pass through six ion channels most affected by drugs in the cell membrane (The six ion channels are as follows: I NaL ; the inward sodium current, I CaL ; the L-type calcium current, I Kr ; the rapidly activating delayed rectifier potassium current, I Ks ; the slowly activating delayed rectifier potassium current, I K1 ; the inward rectifier potassium current, I to ; the transient outward potassium current) during one cycle length (CL), and is computed as equation 2 (Li et al., 2019;Yoo et al., 2021).The mean of qNet values at 1, 2, 3, and 4 multiples of C max or unbound concentration for each drug, the TMS value, showed high accuracy in predicting TdP risks in the prior study of Li et al. (2019).Accordingly, this study calculated the TMS value for CiPA 12 training set drugs and derived logistic regression curves to obtain Threshold 1 and Threshold 2 for classifying TdP risk.Threshold 1 is the TMS value to separate low risk from intermediate/high risk, and Threshold 2 separates high risk from intermediate/low risk.The classifier was based on ordinal logistic regression using the lrm function from version 4.5-0 of the rms package (Chang et al., 2017).To give the logistic   Frontiers in Pharmacology frontiersin.org08 Yoon et al. 10.3389/fphar.2023.1220796

Statistical analysis
Data analysis was performed using pCLAMP (Axon Instruments, Foster City, CA, USA), Origin 2022 (OriginLab Corp, Northampton, MA, USA), Excel (Microsoft, Redmond, WA, USA) and GraphPad Prism (GraphPad Software, San Diego, CA, USA).Statistical significance was determined using Student's t-test.The values were presented as mean ± standard error of the mean (SEM) and were considered statistically significant when the p-value was less than 0.05.Statistically significant differences were assessed using the paired t-test at *p < 0.05, **p < 0.01 or ***p < 0.001.

Multiple effects of drugs on cardiac ion channels
To investigate the effects of the drugs on ion channels involved in cardiac AP formation a whole-cell voltage-clamp technique was employed.In our previous studies, conducted cardiac ionic currents screening using CiPA 12 training set drugs at various multiple of C max (1, 2, 3, and 4).These datasets were then utilized in the optimized ORd model, to predict drug-induced TdP risks with high accuracy.In order to enhance the accuracy of TdP prediction, it was deemed appropriate to perform cardiac ionic currents screening within the same range (1, 2, 3 and 4 multiples of C max or unbound concentration) as TMS.Therefore, in this study, cardiac ionic currents screening was conducted at these specified multiples of C max or unbound concentration to enable more precise prediction of TdP risks.
Considering the C max , the mixture tested decreased the four cardiac ionic currents in a concentration-dependent manner (Figure 5).Additionally, considering the unbound concentration, the mixture tested did not significantly affect the peak Nav1.5, late Nav1.5, Cav1.2, or hERG currents (Figure 6).

TdP risk prediction by TMS
The inhibition rates of four ion channel currents were sampled using the uncertainty quantification algorithm based on the MCMC model to generate 2,000 Hill curves.From each Hill curve, we obtained IC 50 and Hill coefficient (Figure 7).The risk of drug-induced cardiac arrhythmias such as TdP was evaluated using a novel in silico biomarker, qNet, and TMS as proposed by CiPA (Dutta et al., 2017;Li et al., 2019) (Figure 8).As shown in Figure 8A, the two thresholds that classify drugs into three TdP risk categories were also calculated based on the CiPA 12 training set drugs: Threshold 1 has a value of 0.0573 μC/μF and Threshold 2 has a value of 0.0492 μC/μF.When tested considering C max , artesunate and DHA was classified as a low risk of inducing TdP (TMS = 0.064523 μC/μF for artesunate 0.061648 μC/ μF for DHA) (Figure 8B).However, pyronaridine and the mixture were classified as having an intermediate risk of inducing TdP (TMS was 0.051608 μC/μF for pyronaridine and 0.055643 μC/μF for the mixture) (Figure 8B).In contrast, considering unbound concentration, both pyronaridine and mixture were classified as having a low risk of inducing TdP with 0.067301 μC/μF and 0.086473 μC/μF of TMS, respectively.Despite the utilization of the same drug, it was confirmed that the case tested considering the unbound concentration had a lower risk of inducing TdP than the case tested considering C max .

Effects of drugs on cardiac APs in ventricular hiPSC-CMs
To confirm the integrated electrophysiological effects of the drugs, we measured the spontaneous beating of hiPSC-CMs using the whole-cell current-clamp method.Under control conditions, the values of the AP parameters were as follows: −63.1 ± 1.6 mV for MDP, 45.8 ± 4.1 mV/ms for Vmax, 287.0 ± 17.9 ms for APD 90 , 236.1 ± 12.5 ms for APD 50 , 1.2 ± 0.03 for APD 90 /APD 50 ratio and 113.8 ± 1.1 mV for APA (n = 9, mean ± SEM).Previous studies using hiPSC-CMs have demonstrated that ventricular-type cells exhibit certain characteristics, including a plateau phase, APD 90 > 150 ms, and APD 90 /APD 50 ratio of 1.07~1.31(Peng et al., 2010;Honda et al., 2011).The AP waveforms and parameters in our research were consistent with those reported in previous studies.For the analysis and pharmacological testing, only ventricular-type cells were included.The effects of the drugs on AP parameters were normalized with respect to the control value within each individual cell.These normalized values were then presented as bar graphs, depicting the changes in AP shape induced by the drugs.Pyronaridine was tested at four different concentrations, considering its unbound concentration.Applying 1,320 nM pyronaridine induced significant changes in APs; APD 50 and APD 90 increased by 16.07% and 22.34%, respectively, compared to the control (n = 3) (Figure 9).Additionally, pyronaridine significantly decreased the APA of hiPSC-CMs in a concentration-dependent manner.There were no significant changes in the Vmax or MDP with pyronaridine.Artesunate was tested at four different concentrations considering the C max .As shown in Figure 10, artesunate did not significantly alter the cardiac APs of hiPSC-CMs (n = 3).DHA was tested at four different concentrations considering the C max .As shown Figure 11, DHA significantly prolonged the APD 50 of hiPSC-CMs at 4,200 and 126,000 nM by 10.26% and 13.83%, respectively.Additionally, DHA induced significant changes in APD 90 at 4,200, 42,000, 126,000, and 420,000 nM by 7.88, 8.02, 10.74, and 10.16%, respectively.However, it did not significant changes in the Vmax, MDP, or APA.Additionally, the mixture was tested at four different concentrations considering the unbound concentration.It significantly prolonged the APD 50 of hiPSC-CMs at Conc.2, 3, and 4 by 2.99, 7.56, and 8.32%, respectively, but did not change any other AP parameters (MDP, Vmax, APA, or APD 90 ) (Figure 12).

Discussion
In our study, we utilized the CiPA initiative, a novel approach for assessing the proarrhythmic risk of drugs, to evaluate the cardiotoxicity of pyronaridine and artesunate, which were repurposed for COVID-19 treatment during the pandemic.Antimalarial drugs are frequently combined with other drugs to increase their efficacy, reduce the risk of resistant parasites, and shorten the duration of malaria treatment (Kremsner and Krishna,FIGURE 11 Effects of DHA on action potential (AP) parameters of hiPSC-CMs.(A) Typical AP traces of hiPSC-CMs in the control and after exposure to 4,200, 42,000, 126,000 and 420,000 nM DHA.(B-F) Bar graphs illustrating mean normalized values for AP parameters of hiPSC-CMs, including MDP, maximal diastolic potential; APA, AP amplitude; Vmax, maximum rate of depolarization during the upstroke of the AP; APD 50 and APD 90 , AP duration at 50% and 90%.Data are presented as mean ± standard error of the mean (SEM) and analyzed using Student's t-test.*p < 0.05, **p < 0.01 or ***p < 0.001 compared to control (n = 3).

2004
).Among the various antimalarial drug combination therapies, artemisinin-based combination treatments (ACTs) are generally considered the most effective, offering high therapeutic efficacy and real-life safety against malaria (Dini et al., 2018).ACTs consist of an artemisinin derivative paired with a partner drug, such as pyronaridine-artesunate.It has been reported that the combination of pyronaridine and artesunate exhibits a stronger antiviral effect (Nosten and White, 2007;Kurth et al., 2011;Lane et al., 2019).These drugs have been repurposed for COVID-19 treatment, making it crucial to evaluate their cardiovascular adverse reactions, and drug-induced cardiotoxicity.The CiPA initiative proves to be an ideal method for evaluating cardiac safety during drug screening.In our study, we evaluated the drug-induced effects on cardiac arrhythmia for each drug individually and in combination, employing the CiPA initiative.Artesunate was evaluated at four different concentrations considering the C max and was classified as having a low risk of inducing TdP (Figure 8B).Moreover, it had no significant impact on the cardiac APs of hiPSC-CMs (Figure 10).According to Lakhan et al. (2013), intravenous administration of artesunate did not affect the QTc interval, and no significant cardiovascular effects were observed in patients with falciparum malaria.Another study found that intravenous artesunate did not prolong the QT interval, even at high doses (Maude et al., 2009).DHA is the active metabolite of artemisinin compounds (e.g., artemotil, artesunate, artemether) (Woo et al., 1998).DHA was classified as a low TdP risk (Figure 8B).It significantly inhibited cardiac ion channels, but modestly prolonged at APD 90 (10.16%) at 100 times the C max .According to Borsini et al. (2012), the IC 50 on hERG current at 0.1 Hz for DHA is 7.7 ± 0.9 μM; however, it did not significantly prolong the APD 90 .Compared to our research, DHA appears that it significantly interacts with other cardiac ion channels.Furthermore, there have been no adverse cardiovascular effects reported in malaria patients treated with these drugs in clinical trials (White, 2007).Hence, artesunate and DHA have little cardiotoxicity, as indicated in previous reports.
We investigated and compared the effects of pyronaridine and the combination of both drugs on cardiac ion channels and APs at various concentrations based on the C max and unbound concentration.Drugs can bind partially to various plasma or tissue proteins, and typically, it is the unbound (free) fraction of the drug that exerts pharmacological effects (Cervelli and Russ, 2010).Conversely, some drugs tend to accumulate in the myocardium and cells, prolonging their drug action (Tylutki and Polak, 2015).The heart serves as a central pharmacokinetic compartment that is easily accessible to drugs and can also be a site for their side effects (The European Medicines Agency, 2012).Therefore, it is essential to assess drugs at various concentrations, including the free fractions and higher active concentrations.As shown in Figure 8, pyronaridine and its mixtures evaluated based on C max were classified as having intermediate TdP risk.However, when considering unbound concentrations, were classified as having a low risk of inducing TdP.We assessed pyronaridine and mixture at four different concentrations considering the unbound concentration for AP recordings.These significantly inhibited hERG, Nav1.5, and Cav1.2 currents, but remained relatively modest or non-existent effects on APD 90 in hiPSC-CM (Figures 9,12).It is expected that multiple interactions with cardiac ion channels.Similarly, verapamil is well-known hERG blocker but does not prolong the QT interval in clinical settings due to its multichannel effects (Honda et al., 2011).While the potential extrapolation to human is not elucidated, it has been observed that pyronaridine and its metabolites are distributed in high concentrations in the liver, spleen, adrenal gland, kidney, and thyroid gland of rats (The European Medicines Agency, 2012).However, there have been no reports of high concentrations accumulating in the heart.According to the European Medicines Agency (2012), the active concentration of pyronaridine available in the systemic circulation is only 12 ng/mL, as a result of plasma protein or tissue binding.Therefore, when considering electrophysiological recordings, TdP risk prediction using TMS, and pharmacological contexts, it is expected that pyronaridine and the mixture will pose a low proarrhythmogenic risk at supratherapeutic (up to 4 times) free C max .
This study is the first to report the potential cardiotoxicity in electrophysiological terms of pyronaridine and artesunate, which have been repurposed for COVID-19 treatment under the new CiPA initiative (involving multiple ion channel screening, hiPSC-CMs, and a new in silico human-based prediction model that has finished the training phase using CiPA 12 training set drugs).
In summary, this study suggests that pyronaridine, artesunate, and their combination, repurposed for COVID-19 treatment, pose a low risk of adverse cardiac events and cardiotoxicity.Furthermore, the utilization of the CiPA initiative in drug safety assessment platforms can help identify potential cardiotoxicity and serve as an alternative to current in vitro and in vivo systems.

FIGURE 7
FIGURE 7Hill curve for dosage-to-response for pyronaridine on I Na (A), pyronaridine on I NaL (B), pyronaridine on I CaL (C), pyronaridine on I Kr (D).Experimental data points for each cell by circles.95% confidence interval of the fitting on the gray shaded area.Within each panel, the solid red line was fitted using the best parameter of the half-maximal inhibitory concentration (IC 50 ) and Hill coefficient from the nonlinear least square method, whereas the solid black line was the median value from Markov chain Monte Carlo (MCMC) sampling.

FIGURE 8
FIGURE 8 Distribution of the torsade metric score (TMS) for drugs using the optimized ORd model.Drugs are sorted based on the mean of TMS in each dataset.(A) Threshold 1 and threshold 2 were calculated using logistic regression to classify the TdP risk categories (red for high risk, blue for intermediate risk, green for low risk) using CiPA 12 training set drugs.Threshold 1 (separating low from intermediate/high risk) has a value of 0.0573 μC/μF, and threshold 2 (separating high from intermediate/low risk) has a value of 0.0492 μC/μF.(B) The drugs tested considering C max are represented by solid black circle clusters, while drugs tested considering unbound concentration are represented by solid black triangle clusters.

FIGURE 9
FIGURE 9Effects of pyronaridine on action potential (AP) parameters of hiPSC-CMs.(A) Typical AP traces of hiPSC-CMs in the control and after exposure to 13.2, 132, 396, and 1,320 nM pyronaridine.(B-F) Bar graphs illustrating mean normalized values for AP parameters of hiPSC-CMs, including MDP, maximal diastolic potential; APA, AP amplitude; Vmax, maximum rate of depolarization during the upstroke of the AP; APD 50 and APD 90 , AP duration at 50% and 90%.Data are presented as mean ± standard error of the mean (SEM) and analyzed using Student's t-test.*p < 0.05, **p < 0.01 or ***p < 0.001 compared to control (n = 3).

FIGURE 10
FIGURE 10Effects of artesunate on action potential (AP) parameters of hiPSC-CMs.(A) Typical AP traces of hiPSC-CMs in the control and after exposure to 780, 7,800, 23,400, and 78,000 nM artesunate.(B-F) Bar graphs illustrating mean normalized values for AP parameters of hiPSC-CMs, including MDP, maximal diastolic potential; APA, AP amplitude; Vmax, maximum rate of depolarization during the upstroke of the AP; APD 50 and APD 90 , AP duration at 50% and 90%.Data are presented as mean ± standard error of the mean (SEM) and analyzed using Student's t-test.*p < 0.05, **p < 0.01 or ***p < 0.001 compared to control (n = 3).

FIGURE 12
FIGURE 12 Effects of the mixture on action potential (AP) parameters of hiPSC-CMs.(A) Typical AP traces of hiPSC-CMs in the control and after exposure to 13.2, 132, 396, and 1,320 nM pyronaridine and 10.4, 104, 312, and 1,040 nM artesunate.(B-F) Bar graphs illustrating mean normalized values for AP parameters of hiPSC-CMs, including MDP, maximal diastolic potential; APA, AP amplitude; Vmax, maximum rate of depolarization during the upstroke of the AP; APD 50 and APD 90 , AP duration at 50% and 90%.Data are presented as mean ± standard error of the mean (SEM) and analyzed using Student's t-test.*p < 0.05, **p < 0.01 or ***p < 0.001 compared to control (n = 3).P, pyronaridine; A, artesunate.