In Silico Characterization of Repolarization Duration and Variability in the LQT1 Syndrome Under β-Adrenergic Stimulation

Abstract

Enhanced temporal repolarization variability facilitates ventricular arrhythmias in the long QT 1 (LQT1) syndrome, particularly under β-adrenergic stimulation (β-AS). The underlying mechanisms are, however, not fully elucidated. In silico investigation of such mechanisms first requires methods able to reproduce the experimental observations. Here, we describe a method for identification of in silico action potential (AP) models from input voltage traces and we apply it to investigate repolarization variability in LQT1. A combination of Double Greedy Dimension Reduction (DGDR) Unscented Kalman Filter (UKF) was used to estimate the ionic conductances and phosphorylation levels of coupled AP and β-AS models. Over synthetic AP traces from an experimentally-calibrated population of LQT1 cells, combined DGDR-UKF accurately estimated the model parameters, with reduced estimation uncertainty and convergence time. Importantly, combined DGDR-UKF was able to reliably replicate the statistical distributions of AP duration and short-term variability, both at baseline and under β-AS, with relative errors below 4%. Arrhythmogenic AP alternans were reproduced too. In conclusion, our method allows characterization of repolarization duration and variability in LQT1, which is expected to help disentangling the mechanisms underlying adrenergic-induced arrhythmias in this syndrome.