A novel computational model of swine ventricular myocyte reveals new insights into disease mechanisms and therapeutic approaches in Timothy Syndrome

Figures

• 

  Fig. 1. Development and validation of a novel swine model of ventricular action potential. (A) Comparison of the experimental (black, mean±SEM) and fitted (blue) IV curves for INa, ICa, IK1, Ito2, IKr and IKs. INa is presented as normalized current because it was recorded at room temperature, differently from the other currents which were recorded at physiological temperature. Therefore, the INa gating variables had to be temperature-corrected (see Methods). (B) Comparison of simulated (blue) and experimentally measured (black) action potentials (APs) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+ transients (CaTs) during steady-state pacing at 1 Hz. (C) Biomarkers computed on simulated spontaneous APs and CaTs (blue) and their comparison with the experimental values (black, mean±SD). Experimental data derived from Porta-Sanchez et al.4, with the exception of novel Ito2 current recordings.

• 

  Fig. 2. Rate dependence of currents at steady state in the novel swine model of ventricular action potential. Rate dependence of currents at steady state. Black arrows indicate cycle length decrease (rate increase). Top Row: Simulated APs, repeated in each column for timing purposes. Lower Rows (left to right, top to bottom): INa, peak INa detailed time course, late INa, Ito2, ICaL, ICaL increasing peaks with increasing pacing rate, IKr, IKs, IK1, INaCa,i (cytosol), INaCa,ss (subsarcolemmal), and INaK.

• 

  Fig. 3. Sensitivity Analysis conducted in the new swine model and comparison with the human ORd parent model. (A) Schematic illustration of the approach adopted to perform parameter sensitivity analysis. The input matrix X contains randomly generated scaling factors (represented by circles) used to perturb the values of selected parameters in the model. For each simulation, AP and CaT features (squares) are estimated at steady state and their values collected in the matrix Y. Multivariable regression analysis between the input matrix X and the output matrix Y is performed to assess the sensitivity of AP and CaT features to changes in model parameters. The result of this process is the regression matrix B, which coefficients (triangles) quantify the sensitivity of model features to parameter perturbations. (B) APs at the steady state at 1 Hz for the 500 randomly generated models (blue), along with the final selected model (green). (C) Biomarkers computed on simulated APs and CaTs (blue) for the 500 randomly generated models (blue, mean±SD) and their comparison with experimental values (gray, mean±SD) and with the final selected model (green). (D) Colour coded map of sensitivity matrix B…

• 

  Fig. 4. TS1 swine model of ventricular action potential and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+ handling. (A) Comparison between experimental (left) and simulated (right) WT and TS1 APs and CaTs at 1 Hz pacing frequency. Membrane voltage is plotted on the left axis, while [\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+]i (or fluorescence for experimental data) is plotted on the right axis. (B) Comparison between experimental (squares, mean±SD) and simulated (line) APD90 rate adaptation (left) and maximum dV/dt (right) for WT (blue) and TS1 (red). (C) Sensitivity analysis (as in Figure 3). Colour-coded sensitivity matrix B obtained from the WT swine model (left) and the TS1 swine model (right). Experimental data from Porta-Sanchez et al.4.

• 

  Fig. 5. Simulated action potential-clamp in WT and TS1 ventricular models. (A) Simulations of TS1 cell subjected to TS1 AP at 0.5 Hz (left, i.e. “original TS1 cell”) and TS1 cell subjected to WT AP at 0.5 Hz (right). (B) Simulations of WT cell subjected to TS1 AP at 0.5 Hz (left) and WT cell subjected to WT AP at 0.5 Hz (i.e. “original WT cell”). From top to bottom: APs, LTCC Current, CaT and jSR [\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+]. (C) Quantification of jSR \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+ and CaT amplitude in the different situations. Absolute WT values: jSR \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs}…

• 

  Fig. 6. Simulated gene-silencing of LTCC in the TS1 ventricular model. (A) Top: schematic illustration of the simulation design. Simulations were conducted by varying the proportion of silenced TS1-LTCCs (x-axis) and silenced WT-LTCCs (y-axis). Color code represents different degrees of therapeutic efficacy, with red indicating subtherapeutic (ineffective), blue indicating effective and yellow indicating overcorrection of the phenotype. Green box is used to represent values closest to WT. Bottom: Simulations results at 0.5 Hz (left), 1 Hz (center) and 2 Hz (right) for the three evaluated phenotypes: APD90 (top row), maximum upstroke velocity (middle row), and jSR [\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+]. (B) Comparison of the effects between the most-balanced gene therapy dosing (75% TS1-LTCCs and 25% WT-LTCCs silenced) (top) vs. non-specific targeting with LTCC blockers (bottom) on repolarization duration, depolarization reserve and cellular…

• 

  Fig. 7. Role of secondary contributors to the TS1 cellular phenotype. (A) Contribution of CaMKII activation to TS1 phenotype assessed by simulating the effects of CaMKII inhibition (green) and comparing with TS1 (red) and WT (blue) on APD90 (left), upstroke velocity (center) and cellular \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {Ca}^{2+}$$\end{document}Ca2+ overload (right). (B) Contribution of altered repolarizing currents (40% IKs reduction and a +20.7 mV shift with 5% reduced conductance in the current-voltage relationship of Ito2) to the TS1 cellular phenotype, simulated by restoring WT repolarizing currents in TS1 (green) vs. TS1 (red) and WT (blue). (C) Joint contribution of CaMKII activation and altered repolarizing currents to shaping the TS1 phenotypes (green) vs. TS1 (red) and WT (blue). (D) Simulated effects of increasing IKr current with ICA-105574 on the TS1 cellular phenotype. Details for the estimation of parameters for ICA-105574 simulation are presented in Supplementary Figure 9.…

• 

  Fig. 8. Arrhythmogenic Potential of TS1 Cardiomyocytes. (A) Left: experimental results showing increased expression of exon 8A in TS1 as compared to WT. Experimentally-derived average proportion of G406R-mutated LTCCs was 22%. Right: simulated relationship between increasing proportion of G406R-mutated LTCCs and APD90 at 0.5 Hz (red), 1 Hz (blue) and 2 Hz (black). Inset shows the behavior at 0.5 Hz when the proportion of G406R LTCCs exceeds 26%, characterized by appearance of EADs (starting at 26.6%) and 2:1 capture (starting at 27%). Green bar depicts the experimentally-observed range of G406R-mutated LTCCs. (B) Simulated traces of TS1 APs and CaTs at 0.5 Hz when introducing slowing of inactivation of the hl gate of INaL (tau-hl = 600 ms). (C) Simulated traces of TS1 APs and CaTs at 0.5 Hz when introducing 0.925 mV left-shift in Cav1.2 activation. (D) Simulated traces of TS1 APs and CaTs at 0.5 Hz when introducing mathematical imposition of cytosolic \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox…