Size matters, proportion too: coupling of experiments and theory reveals relative roles of K⁺ channels in action potential stability

Cardiac repolarization is a delicate and highly regulated process, governed by a complex interplay of ionic currents through channels and transporters, as well neurohormonal modulators. While drugs that prolong the cardiac action potential are particularly effective in preventing re-entrant arrhythmias, action potential duration (APD)-prolonging drugs, such as K+ channel blockers, can also destabilize repolarization, causing early afterdepolarizations (EADs) and potentially triggering lethal ventricular tachyarrhythmias, i.e. torsades de pointes (TdP). Despite tremendous advances in experimental and theoretical approaches and insight in cardiac physiopathology, we are still trying to identify key biophysical details that are important for fully understanding the mechanistic underpinning of cardiac repolarization and its abnormalities. One element that remains not fully resolved is the functional significance of the slow (vs. rapid) component of the delayed rectifier K+ current (IKs vs. IKr). IKs is sensitive to adrenergic stimulation and internal [Ca2+] (Bartos et al. 2017), but under basal conditions its density has typically been measured to be much lower than the rapid component IKr in large mammals, including humans (Jost et al. 2007). Nevertheless, upon β-adrenergic receptor (β-AR) stimulation IKs increases (more than IKr does) and counteracts ICaL enhancement to prevent potentially deleterious repolarization delays. Indeed, arrhythmic events in congenital long QT syndromes linked to IKs dysfunction are typically triggered by exercise and stress, and can be prevented by β-AR blockade. IKs is also a critical contributor to repolarization reserve, whereby IKs becomes increasingly important in limiting excessive APD prolongation when other repolarizing currents are suppressed (Jost et al. 2005). In a white paper in this Special Issue of The Journal of Physiology, we highlight several technical limitations that undermine the experimental identification of the precise role of each repolarizing current to shaping AP repolarization (Grandi et al. 2017). These include lack of drug selectivity, lack of critical channel regulatory components in heterologous expression systems, current rundown, or digestion artefacts during cardiac myocyte isolation. Furthermore, while iteration of mathematical models and experiments has been used for decades to quantitatively study cardiac myocyte physiology, challenges still exist for mathematical modelling, including the need of coherent experimental datasets for model parameterization, as opposed to the use of heterogeneous data from multiple sources, species and conditions. Also, only in a few cases is the iterative loop actually closed, whereby experiments are utilized to inform model development, and novel insight from simulations are fed back into new experimental protocols to test model predictions. In this issue of The Journal of Physiology, some of these challenges have been elegantly addressed by Devenyi, Ortega and colleagues, who devised a powerful and efficient combination of computational and experimental techniques that led to improvement of an existing computational model of the guinea pig ventricular myocyte and new mechanistic insight in the regulation of APD in this species (Devenyi et al. 2017). A key strength of this study is the clever approach, which affords numerous advantages over conventional electrophysiological and modelling methodologies. The dynamic clamp technique allows highly selective control over individual system parameters, and rapid experimental screening of computationally based hypotheses. Information-rich dynamic clamp protocols and population-based computational approaches allowed acquiring comprehensive datasets efficiently. This afforded comparison of the measured and simulated cell responses to perturbations in many parameters under a uniform set of conditions and in the same species. Dynamic clamp was used here recursively, in combination with genetic algorithms, to optimize and validate an existing computational model of the guinea pig ventricular myocyte. Interestingly, this unbiased model adjustment consistently predicted an increase in IKr and a drastic decrease (to ∼4%) in IKs compared to the baseline model. The prediction was validated experimentally, and revealed modest APD prolongation upon IKs block in both simulation and experiment. This led to the hypothesis that the balance or proportion of these repolarizing currents, more than their magnitude, influences the system response to perturbations. Indeed, subsequent simulations with the adjusted model showed that increasing the IKs/IKr ratio, without changing the overall duration of repolarization, rendered the cell more resistant to pro-arrhythmic stimuli, suggesting the intriguing notion that IKs can stabilize APD and suppress EADs in response to perturbations better than IKr. Future experimental studies are warranted to validate the role of the IKs/IKr ratio on APD stability and EAD formation. Devenyi et al. (2017) suggest that the greater capability of IKs to stabilize the APD is owing to its slow activation and rapid deactivation, which favour IKs increase with APD prolongation, in contrast to faster activation and inactivation of IKr. Notably, increasing [Ca2+]i and β-AR stimulation were shown to produce a similar spectrum of biophysical consequences on IKs (Bartos et al. 2017), resulting in increased current density and slowed deactivation kinetics. Thus, both β-AR stimulation and high [Ca2+]i may work in concert to stabilize APD against AP-prolongation stimuli by both increasing the GKs/GKr ratio and enhancing IKs during prolonged repolarization (and shorter diastolic intervals). Many IKs studies, including that of Devenyi et al. (2017), have utilized guinea pig as a model system and taken advantage of the large current density in this species. However, it would be interesting to quantify the role of the IKs/IKr ratio in APD stability in various species, given well-known species differences in overall repolarization dynamics, IKs magnitude and kinetics (e.g. faster activation and deactivation in human vs. guinea pig myocytes) (Jost et al. 2007). While several questions remain, this work of Devenyi et al. (2017) beautifully showcases the power of iterative coupling of simulations and experiments to both further refine and improve the models and computational tools for use by many and advance our understanding of pathophysiologically relevant mechanisms. None. This work was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health (R01HL131517) and American Heart Association (15SDG24910015).