The normal functioning of the heart resides in its coordinated and effective pumping capacity. Contraction of the heart is provided by the ordered and intricate design of the contractile apparatus, the sarcomere, while the coordination of contraction is accomplished by the cardiac conduction system.1 At the cellular level, the electrical properties of the contracting cardiomyocytes are governed by a complex array of inward and outward ion currents that configure the cardiac action potential. If we take as a paradigm the human cardiomyocyte, upstroke sodium current (INa) is responsible for depolarization of the cardiomyocyte plasma membrane,2 whereas a variety of outward potassium currents (IKur,IKr, IKs, IK1,ITO) provide the means to repolarize the cell membrane as well as to keep constant the resting membrane potential.3,4 Electrical coupling—that is to say from conduction to contraction—is provided mainly by calcium currents (ICa).5
It is important to acknowledge the nature of the regulation of the cardiac action potential, since limited variability is observed within the sodium currents in atrial and ventricular cardiomyocytes, a situation that similarly applies to the electrical coupling governed by calcium currents. However, a large variability is observed for the repolarization and resting membrane potential.6,7 For example, distinct resting membrane potential and cardiac action potential morphologies are observed for atrial and ventricular myocytes, and even within ventricular myocytes a differential contribution of the repolarization currents leads to distinct cardiac action potential morphology in epicardial, mid-myocardial, and endocardial myocytes.8 Such variability is at the soul of the contribution by Danielsson et al.9 in this issue, since they provide seminal evidence of the large repolarization variability and its influence on cardiac toxicity during embryogenesis. Moreover, the distinct sodium and calcium currents across species, such as in the most frequently used experimental models (mouse, rat, dogs, and pigs), contribute similarly and equally to the configuration of the cardiac action potential, while a large variability is observed for the repolarization currents. As an example, IKr and IKs contribute largely to the plateau phase of cardiac repolarization in human cardiomyocytes, while in rats and mice, the contribution of these currents is limited, with the repolarization phase depending mainly on the configuration of IKur and ITO.10 Such complexity is directly applicable to the underlying molecular substrates. In this way, IK currents are generated by channels formed by distinct pore-forming subunits (Kcnq1 and Kcnh2), which can interact between themselves as well as with a large number of ancillary subunits11–13 configuring IKr and IKs, respectively. Similarly, ITO is differentially modulated by several pore-forming subunits (Kv4.2, Kv4.3), which play distinct roles in distinct species.14,15 Thus, intuitively, we can foresee that repolarization is a pivotal step in the cardiac action potential control, given its tight and complex regulation. In this setting, alterations in repolarization will thus have a major impact on the electrical behaviour of the heart as revealed by point mutations in a large number of repolarization ion channels that are associated with arrhythmogenic life-threatening syndromes such as Brugada and long QT.16 Importantly, the large diversity and complex functionality across species has limited in many cases the use of genetically amenable experimental models such as mice as a model organism for the extrapolation of cardiac electrical properties of the heart to humans.10 The use of larger animal models, such as dogs and pigs, while more closely resembling the ion current composition of the cardiac action potential of humans, has provided critical but yet limited insights, probably due to hindrances such as the lack of large-scale genetic manipulation, inherent longer gestational periods, and higher experimental costs.
Significantly, apart from genetically determined, repolarization-associated cardiac arrhythmias, it is widely acknowledged that a variety of drugs can elicit repolarization defects that lead to equally life-threatening cardiac arrhythmias such as ventricular fibrillation in adulthood, which are highly prevalent in Brugada and long QT syndromes.16 Awareness of these potential threats is continuously updated in dedicated open-access online resources such as QTdrugs (www.qtdrugs.org) and BrugadaDrugs (www.brugadadrugs.org). However, a major challenge remains concerning the current usage of drugs with long QT risk during pregnancy given our limited knowledge on the differential contribution of distinct ion currents to the configuration of the gestational cardiac action potential as well as on the significance of electric developmental insults that might lead to cardiac arrhythmias in adulthood. In an effort to minimize the side effects of novel drugs, regulatory agencies require the testing of putative side effects on cardiac repolarization in distinct experimental animal models. However, this poses a major challenge given the difficulty of directly extrapolating animal model findings to the human situation; this is particularly true regarding pregnancy risks as there is limited information available about putative teratogenic effects of drugs during gestation in experimental models and their translation to humans.
While there is a large body of evidence demonstrating that ion channel expression is largely heterogeneous and dynamic during cardiac development, including sodium, calcium, and most importantly potassium channels,17–20 there is little information in the literature regarding their functional behaviour and/or implications for impairment. The study of Danielsson et al.9 brings novel and highly relevant insights on this front. By comparing the IK current contribution to the cardiac action potential within three distinct species—rat, rabbit, and human—at distinct developmental stages, they provide information on the species-specific gestational differences. Molecular analyses of the genetic determinants of these currents further underscore their findings: IKr and IKs are highly stable in human cardiomyocytes during gestation, IKr is highly stable but IKs is absent in rabbits, whereas IKr is only present during the early gestational period of rat development and is basically absent in foetal stages, when Iks peaks. Importantly, these observations nicely correlate with the increased developmental toxicity previously reported in different species. Furthermore, pharmacological treatment with an IKr blocker (E-4031) demonstrates the pivotal role of IKr, since prolongation of the cardiac action potential was observed in those cells containing high levels of Kcnh2 across species. ECG tracings in treated rat embryos further underscored the provoked cellular electrophysiological impairment, since prolonged QT intervals were also observed. Thus, these data have a high degree of relevance to the understanding of the biological insults of long QT-prone drugs currently used during pregnancy that might result in unexpected and undesired abortions. They also highlight the critical role of IKr as a common mechanism leading to severe life-threatening defects during embryogenesis. Equally important, the study of Danielsson et al.9 provides seminal findings regarding the adequate usage and limitations of using rats and/or rabbits as experimental models or surrogates for humans.
There are new avenues of research opening up ahead of us that might solve our current concerns. Two novel experimental models for arrhythmogenic drug testing have been recently put forward. The first is the zebrafish, which has been proposed as a novel surrogate since there have been major achievements in cardiac electrophysiological characterization and high-throughout platforms have been developed.21,22 Although the action potential and electrocardiogram morphology quite nicely resemble those of humans, other caveats such as limited biophysical characterization of the cardiac currents and overt distinct cardiac morphology remain a challenge. The second new platform for individualized drug testing is patient-derived induced pluripotent stem cells (iPS), as reported in recent seminal works with long QT syndrome patient-derived iPS23,24; however, a major limitation yet remains to obtain fully differentiated iPS-derived cardiomyocytes.
Conflict of interest: none declared.