Jonathan D. Moreno

A Computational Model to Predict the Effects of Class I Anti-Arrhythmic Drugs on Ventricular Rhythms

Two- and three-dimensional models of cardiac excitability based on sodium channel kinetics can predict the adverse effects of class I anti-arrhythmic drugs. Crowdsourcing the Heart for Drug Screening The old way: Consult a specialist to answer your question. The new way: Consult a crowd of generalists who in the aggregate can come up with a better answer. The old way—testing drugs on single cardiac cells in vitro—has not worked well for screening out potential anti-arrhythmia agents that can occasionally block conduction in the heart or exacerbate arrhythmia, serious problems that cause sudden death in treated patients. Instead, Moreno et al. have called on the crowd by building a model of heart tissue that includes many cardiac cells and their interactions. When anti-arrhythmia drugs are “applied” to the model’s beating heart tissue—but not when they are applied to the single cardiac cells that make up the model—the drugs that cause side effects, and the concentrations at which they do so, are revealed, results that the authors were able to validate experimentally. The model starts with the detailed kinetics of the heart’s sodium channels, first in the context of a single cell, then in two- and three-dimensional cardiac tissue. The authors compared the action of lidocaine, a class 1B anti-arrhythmic drug not known to cause conduction block, and flecainide, a prototypical class 1C drug that carries a warning from the Food and Drug Administration. In the modeled analyses of single cardiac cells, both drugs slowed excitability at concentrations that matched those used in patients, but the cells retained the ability to generate action potentials. But when the model incorporated coupled groups of cells, the behavior of the drugs diverged. Lidocaine lowered excitability without causing block, but at the higher concentrations (used clinically), flecainide caused serious conduction block when heart rates reached 160 beats per minute. Experiments in rabbit heart confirmed the results of the model. In scaled-up, 500 by 500 groups of cells, the authors’ model could also successfully predict the tendency of flecainide, but not lidocaine, to make the heart extra sensitive to heartbeats occurring too early or too late, an effect that causes even more severe arrhythmias in patients when they take anti-arrhythmia drugs. Again, experiments in rabbit hearts replicated the model’s predictions, as did simulations of anatomically accurate human hearts derived from magnetic resonance imaging images. The ability of this sophisticated model of living cardiac tissue to replicate the clinical adverse effects of lidocaine and flecainide is promising, but it will be necessary to validate its performance with other drugs to understand how to deploy it most effectively. Ideally, such models will be useful for screening out potential arrhythmic drugs that promote conduction block or exacerbate arrhythmias. Such a view of how drugs affect the collective activity of cardiac cells should help in these situations in which the cure proves more deadly than the disease. A long-sought, and thus far elusive, goal has been to develop drugs to manage diseases of excitability. One such disease that affects millions each year is cardiac arrhythmia, which occurs when electrical impulses in the heart become disordered, sometimes causing sudden death. Pharmacological management of cardiac arrhythmia has failed because it is not possible to predict how drugs that target cardiac ion channels, and have intrinsically complex dynamic interactions with ion channels, will alter the emergent electrical behavior generated in the heart. Here, we applied a computational model, which was informed and validated by experimental data, that defined key measurable parameters necessary to simulate the interaction kinetics of the anti-arrhythmic drugs flecainide and lidocaine with cardiac sodium channels. We then used the model to predict the effects of these drugs on normal human ventricular cellular and tissue electrical activity in the setting of a common arrhythmia trigger, spontaneous ventricular ectopy. The model forecasts the clinically relevant concentrations at which flecainide and lidocaine exacerbate, rather than ameliorate, arrhythmia. Experiments in rabbit hearts and simulations in human ventricles based on magnetic resonance images validated the model predictions. This computational framework initiates the first steps toward development of a virtual drug-screening system that models drug-channel interactions and predicts the effects of drugs on emergent electrical activity in the heart.

Pathophysiology of the cardiac late Na current and its potential as a drug target

A pathological increase in the late component of the cardiac Na(+) current, I(NaL), has been linked to disease manifestation in inherited and acquired cardiac diseases including the long QT variant 3 (LQT3) syndrome and heart failure. Disruption in I(NaL) leads to action potential prolongation, disruption of normal cellular repolarization, development of arrhythmia triggers, and propensity to ventricular arrhythmia. Attempts to treat arrhythmogenic sequelae from inherited and acquired syndromes pharmacologically with common Na(+) channel blockers (e.g. flecainide, lidocaine, and amiodarone) have been largely unsuccessful. This is due to drug toxicity and the failure of most current drugs to discriminate between the peak current component, chiefly responsible for single cell excitability and propagation in coupled tissue, and the late component (I(NaL)) of the Na(+) current. Although small in magnitude as compared to the peak Na(+) current (~1-3%), I(NaL) alters action potential properties and increases Na(+) loading in cardiac cells. With the increasing recognition that multiple cardiac pathological conditions share phenotypic manifestations of I(NaL) upregulation, there has been renewed interest in specific pharmacological inhibition of I(Na). The novel antianginal agent ranolazine, which shows a marked selectivity for late versus peak Na(+) current, may represent a novel drug archetype for targeted reduction of I(NaL). This article aims to review common pathophysiological mechanisms leading to enhanced I(NaL) in LQT3 and heart failure as prototypical disease conditions. Also reviewed are promising therapeutic strategies tailored to alter the molecular mechanisms underlying I(Na) mediated arrhythmia triggers.

Computational approaches to understand cardiac electrophysiology and arrhythmias

Cardiac rhythms arise from electrical activity generated by precisely timed opening and closing of ion channels in individual cardiac myocytes. These impulses spread throughout the cardiac muscle to manifest as electrical waves in the whole heart. Regularity of electrical waves is critically important since they signal the heart muscle to contract, driving the primary function of the heart to act as a pump and deliver blood to the brain and vital organs. When electrical activity goes awry during a cardiac arrhythmia, the pump does not function, the brain does not receive oxygenated blood, and death ensues. For more than 50 years, mathematically based models of cardiac electrical activity have been used to improve understanding of basic mechanisms of normal and abnormal cardiac electrical function. Computer-based modeling approaches to understand cardiac activity are uniquely helpful because they allow for distillation of complex emergent behaviors into the key contributing components underlying them. Here we review the latest advances and novel concepts in the field as they relate to understanding the complex interplay between electrical, mechanical, structural, and genetic mechanisms during arrhythmia development at the level of ion channels, cells, and tissues. We also discuss the latest computational approaches to guiding arrhythmia therapy.

Ranolazine for Congenital and Acquired Late INa-Linked Arrhythmias In Silico Pharmacological Screening

Rationale: The antianginal ranolazine blocks the human ether-a-go-go–related gene–based current IKr at therapeutic concentrations and causes QT interval prolongation. Thus, ranolazine is contraindicated for patients with preexisting long-QT and those with repolarization abnormalities. However, with its preferential targeting of late INa (INaL), patients with disease resulting from increased INaL from inherited defects (eg, long-QT syndrome type 3 or disease-induced electric remodeling (eg, ischemic heart failure) might be exactly the ones to benefit most from the presumed antiarrhythmic properties of ranolazine. Objective: We developed a computational model to predict if therapeutic effects of pharmacological targeting of INaL by ranolazine prevailed over the off-target block of IKr in the setting of inherited long-QT syndrome type 3 and heart failure. Methods and Results: We developed computational models describing the kinetics and the interaction of ranolazine with cardiac Na+ channels in the setting of normal physiology, long-QT syndrome type 3–linked &Dgr;KPQ mutation, and heart failure. We then simulated clinically relevant concentrations of ranolazine and predicted the combined effects of Na+ channel and IKr blockade by both the parent compound ranolazine and its active metabolites, which have shown potent blocking effects in the therapeutically relevant range. Our simulations suggest that ranolazine is effective at normalizing arrhythmia triggers in bradycardia-dependent arrhythmias in long-QT syndrome type 3 as well tachyarrhythmogenic triggers arising from heart failure–induced remodeling. Conclusions: Our model predictions suggest that acute targeting of INaL with ranolazine may be an effective therapeutic strategy in diverse arrhythmia-provoking situations that arise from a common pathway of increased pathological INaL.

In silico prediction of drug therapy in catecholaminergic polymorphic ventricular tachycardia

The mechanism of therapeutic efficacy of flecainide for catecholaminergic polymorphic ventricular tachycardia (CPVT) is unclear. Model predictions suggest that Na+ channel effects are insufficient to explain flecainide efficacy in CPVT. This study represents a first step toward predicting therapeutic mechanisms of drug efficacy in the setting of CPVT and then using these mechanisms to guide modelling and simulation to predict alternative drug therapies.

Parameterization for in-silico modeling of ion channel interactions with drugs

Since the first Hodgkin and Huxley ion channel model was described in the 1950s, there has been an explosion in mathematical models to describe ion channel function. As experimental data has become richer, models have concomitantly been improved to better represent ion channel kinetic processes, although these improvements have generally resulted in more model complexity and an increase in the number of parameters necessary to populate the models. Models have also been developed to explicitly model drug interactions with ion channels. Recent models of drug-channel interactions account for the discrete kinetics of drug interaction with distinct ion channel state conformations, as it has become clear that such interactions underlie complex emergent kinetics such as use-dependent block. Here, we describe an approach for developing a model for ion channel drug interactions. The method describes the process of extracting rate constants from experimental electrophysiological function data to use as initial conditions for the model parameters. We then describe implementation of a parameter optimization method to refine the model rate constants describing ion channel drug kinetics. The algorithm takes advantage of readily available parallel computing tools to speed up the optimization. Finally, we describe some potential applications of the platform including the potential for gaining fundamental mechanistic insights into ion channel function and applications to in silico drug screening and development.

Reply from Pei‐Chi Yang, Jonathan D. Moreno, Mao‐Tsuen Jeng, Xander H. T. Wehrens, Sergei Noskov and Colleen E. Clancy

We appreciate Williams et al. (2016) taking the time to comment on our recently published study (Yang et al. 2016). In their letter, the authors question the ‘usefulness’ of the computational modelling and simulation approaches that we used in part because as they state, ‘The blocking parameters used in Yang et al. (2016) are based on values reported in Hilliard et al. (2010) and subsequent publications from the same group.’ This statement does not reflect the careful process that we actually used in building our modelling approaches, where we rather considered the full range of experimentally measured IC50 values for flecainide interaction that have been reported in multiple studies. In addition to the assumption of IC50 = 0 μM (i.e. no interaction with RyR) as reported by the Williams group (Bannister et al. 2015), we reported the following in our paper (Yang et al. 2016): ‘Isoproterenol-stimulated Ca2+ waves in CASQ2 knockout (KO) CASQ2(−/−) mice were inhibited by flecainide with an IC50 of 2.0 ± 0.2 μM (Hwang et al. 2011), while other experimental preparations measured an IC50 range from 2 to 17 μM (Brunton et al. 2010; Hilliard et al. 2010; Hwang et al. 2011; Mehra et al. 2014) . . . We also predicted cases for variable flecainide IC50 = 3, 4, and 5 μM shown in Fig. 1.’ The model simulations led to the predictions that IC50 values above 5 μM are too low to show therapeutic benefit to normalize the catecholaminergic polymorphic ventricular tachycardia (CPVT) phenotype. An alternative interpretation is that the concentration of flecainide near the receptor is considerably higher than in the bulk water compartments, a possibility supported by our physics-based approach (Fig. 5 in Yang et al. 2016) that shows accumulation of flecainide on the membrane surface and very favourable conditions for neutral flecainide in the hydrophobic core of the membrane. Detailed investigations into membrane partitioning of drugs are ongoing in our group. The point of the simulations in our study was to make predictions about the necessary and sufficient targets of flecainide and the range of IC50 that would allow for normalization of the CPVT phenotype since the experimental literature has shown such variety in reported values. When we started the investigation reported in Yang et al. (2016), we had no preconceived intent or notion about the results. The predictions are the resulting outputs of the model, and suggest that Na+ channel block alone is not sufficient to prevent the CPVT phenotype. The critical point here is that the disparity in sensitivity of the dose–response for flecainide interaction with the RyR depends on the experimental approach being used. This issue has been the subject of discussion by others (Steele et al. 2013; Sikkel et al. 2013b; Smith & MacQuaide, 2015). Williams et al. describe their recent work in their letter. It is important to mention, however, the numerous other studies that report alternative data and explanations. Some in native myocytes show very clear effects of flecainide on spontaneous Ca2+ release (i.e. Ca2+ waves) under experimental conditions where cytosolic [Ca2+] and [Na+] are clamped, demonstrating a direct action of flecainide on RyR2-mediated sarcoplasmic reticulum (SR) Ca2+ release (Savio-Galimberti & Knollmann, 2015; Hilliard et al., 2010; Galimberti & Knollmann, 2011). Moreover, in native myocytes, flecainide does not inhibit physiological Ca2+ current-induced SR Ca2+ release but only inhibits spontaneous SR Ca2+ release, which occurs in the setting of diastolic [Ca2+] (i.e. 100 nM) (Hilliard et al. 2010). Such conditions are difficult to model using RyR2 channels incorporated into artificial bilayers and hence were never tested by the group of Williams et al. Other studies demonstrate a clear benefit of flecainide in the clinical CPVT setting, but not in experiments with other Na+ channel blockers (Watanabe et al. 2009; Hwang et al. 2011; van der Werf et al. 2011). Williams et al. performed single-channel experiments in an experimental model comprising phosphatidylethanolamine (PE) bilayers to show that flecainide does not block ion current by binding to a site within the cytosolic domain of the pore-forming domain of RyR2. However, other data and the physics-based computational approaches in our paper suggest that lipophilic drug access may be critical and is a vital component of drug interactions with membrane protein targets such as RyR2. The potential of mean force calculations we performed in our study suggest that flecainide concentration in the lipid phase could be substantially greater than what would be expected in the bilayer studies. Carvedilol is another example of a very hydrophobic/lipophilic drug that interacts with RyR2 without blocking unitary conductance in single-channel experiments. Liposome partitioning experiments suggest that up to 90% of carvedilol molecules are lipid-phase localized (Cheng et al. 1996). The lipophilic access mechanism would imply different dose–response ratios and use-dependent features of drug interaction with the RyR2 target in contrast to a single-site drug block mechanism endorsed by Williams et al. It is important to point out that lipophilic access mechanisms have been shown recently for various membrane targets found in the heart (Lees-Miller et al. 2015; Boiteux et al. 2014) and are likely to exist for RyR2 given the lipophilicity of many drugs interacting with this channel. Williams et al. have undertaken valuable biophysical studies using purified recombinant channels in artificial lipid bilayers. We argue, however, that such a system is far removed from the physiological reality and cannot unequivocally prove the absence of a flecainide interaction with RYR2 channels in a native cellular environment. For example, Cannon et al. (2003) reconstituted RyR2 into a bilayer composed by 1-palmitoyl-2-oleoylphosphatidylethanolamine (POPE) and 1-palmitoyl-2-oleoyl-phosphatidylcholine (POPC) showing that channel activity depends critically on the bilayer composition. Another study showed that the polyunsaturated fatty acid eicosapentanoic acid (EPA) exerts its antiarrhythmic effect by reducing the opening probability of RyR2 (Swan et al. 2003). This is important, because the artificial bilayer used by Williams et al. was composed of 100% (PE), but the actual SR lipid content from dog hearts showed the presence of triglycerides, cholesterol and other phospholipids like phosphatidylinositol (PI), phosphatidylcholine (PC), sphingomyelin (SM) and phosphatidylserine (PS). Most of these

Mechanisms and models of cardiac sodium channel inactivation

ABSTRACT Shortly after cardiac Na+ channels activate and initiate the action potential, inactivation ensues within milliseconds, attenuating the peak Na+ current, INa, and allowing the cell membrane to repolarize. A very limited number of Na+ channels that do not inactivate carry a persistent INa, or late INa. While late INa is only a small fraction of peak magnitude, it significantly prolongs ventricular action potential duration, which predisposes patients to arrhythmia. Here, we review our current understanding of inactivation mechanisms, their regulation, and how they have been modeled computationally. Based on this body of work, we conclude that inactivation and its connection to late INa would be best modeled with a “feet-on-the-door” approach where multiple channel components participate in determining inactivation and late INa. This model reflects experimental findings showing that perturbation of many channel locations can destabilize inactivation and cause pathological late INa.

Interaction of Resveratrol with Lipid Membranes

Background: Resveratrol is a phytoalexin synthesized by plants. It has antioxidant properties and is a popular nutritional supplement. Beneficial properties of resveratrol, such as, anticancer and anti-inflammatory properties have been reported. Resveratrol is a constituent of red wine and found in the skin of red grapes. In order to understand the interaction between resveratrol and biological membranes, we evaluated the effect of resveratrol on model lipid membranes using differential scanning calorimetry (DSC) and computational studies. Methods: Phospholipids such as, Dilauroylphosphatidylcholine (DLPC), Dimyristoyphosphatidylcholine (DMPC), Dipalmitoylphosphatidyl choline (DPPC), Distearoylphosphatidylcholine (DSPC), and 1-palmitoyl-2-oleyl phosphatidylcholine (POPC) were purchased from Avanti Polar Lipids (Alabaster, Alabama). Each phospholipid was mixed with resveratrol at different molar ratios, phospholipid: resveratrol, 10:1, 10:3 and 10:5. Computational simulations were also performed to evaluate the interactions of DSPC and resveratrol.Results: Resveratrol abolishes only the transition of the DLPC, which is the shortest phospholipid of those tested. It reduces the transition temperature for all the other phospholipids even at the lowest ratio tested, phospholipid: resveratrol, 10:1. Resveratrol reduces the transition temperature of DSPC from 55 °C to 51 °C. Furthermore, using DSC, we also observed another transition, a sharp exothermic peak above 275 °C in the interaction of resveratrol with DSPC. We performed computational simulations of the DSPC membrane at different temperatures with and without resveratrol. The simulation indicates that resveratrol affects the transition temperature of the DSPC, which is in agreement with our DSC data. In conclusion, our data indicate that resveratrol abolishes the transition of DLPC and acts as a plasticizer for phospholipids with longer fatty acyl chains.

Ranolazine for Congenital and Acquired Late INa-Linked Arrhythmias

Rationale: The antianginal ranolazine blocks the human ether-a-go-go–related gene–based current IKr at therapeutic concentrations and causes QT interval prolongation. Thus, ranolazine is contraindicated for patients with preexisting long-QT and those with repolarization abnormalities. However, with its preferential targeting of late INa (INaL), patients with disease resulting from increased INaL from inherited defects (eg, long-QT syndrome type 3 or disease-induced electric remodeling (eg, ischemic heart failure) might be exactly the ones to benefit most from the presumed antiarrhythmic properties of ranolazine. Objective: We developed a computational model to predict if therapeutic effects of pharmacological targeting of INaL by ranolazine prevailed over the off-target block of IKr in the setting of inherited long-QT syndrome type 3 and heart failure. Methods and Results: We developed computational models describing the kinetics and the interaction of ranolazine with cardiac Na+ channels in the setting of normal physiology, long-QT syndrome type 3–linked &Dgr;KPQ mutation, and heart failure. We then simulated clinically relevant concentrations of ranolazine and predicted the combined effects of Na+ channel and IKr blockade by both the parent compound ranolazine and its active metabolites, which have shown potent blocking effects in the therapeutically relevant range. Our simulations suggest that ranolazine is effective at normalizing arrhythmia triggers in bradycardia-dependent arrhythmias in long-QT syndrome type 3 as well tachyarrhythmogenic triggers arising from heart failure–induced remodeling. Conclusions: Our model predictions suggest that acute targeting of INaL with ranolazine may be an effective therapeutic strategy in diverse arrhythmia-provoking situations that arise from a common pathway of increased pathological INaL.