Pro-Arrhythmic Effects of Discontinuous Conduction at the Purkinje Fiber-Ventricle Junction Arising From Heart Failure-Induced Ionic Remodeling – Insights From Computational Modelling

FIGURE 1. (A–D) Simulated time courses of APs in ENDO, M, EPI and PF cell models which were compared to the original Benson et al. model (2008) and Aslanidi et al. model (2009). (E) Simulated time courses of [Ca2+]i during an evoked AP in the M cell model in comparison to experimental data from O’Rourke et al. (O'Rourke et al., 1999). (F) Computed diastolic and systolic [Ca2+]i from the modified models in comparison to the original Benson et al. (2008) models and experimental data (O'Rourke et al., 1999). (G) Computed APDs at 1 Hz in ventricular cell models in comparison to the original Benson et al. (2008) models and experimental data (experimental data are listed in Supplementary Tables S10–S12). (H) Computed APDs at different simulation rates PF cell model in comparison to the original Aslanidi et al. model (2009) model and experimental data (experimental data are listed in Supplementary Table S13).

FIGURE 2. (A) Computed time courses of APs and HF-remodelled ionic currents in the CTL and HF conditions for ventricular and PF cell models. Five major HF-remodelled ionic currents were shown (see Supplementary Text S1). (B) Maximum upstroke velocity. Experimental data (Balati et al., 2000; Maguy et al., 2009; Hegyi et al., 2019) were shown in solid dots with error bars. (C) Computed APD in the CTL and HF conditions for ventricular and PF models. (D) Simulated [Ca2+]i in the CTL and HF conditions. Experimental data (O'Rourke et al., 1999) was also shown in the inset.

FIGURE 3. Top (A–I) Effect of each HF-induced change in ionic current on AP. The APs from the CTL condition were superimposed on those when HF-remodelling on each of the remodelled ionic currents alone was considered. Bottom (J–R) Effect of excluding each of the HF-induced changes in ionic currents from HF-remodelling on the ventricular APs. The APs in the HF condition (black solid line; all HF-induced changes were considered) was superimposed on those when HF-induced changes in an individual ionic current (black dash line) was excluded.

FIGURE 4. (A–D) AP propagation along the 1D model of ventricle (A,B) and PF-ventricle (C,D) in both CTL and HF conditions. This PF strand had a length of 18 mm and the ventricular strand had a length of 15 mm (Akar and Rosenbaum, 2003; Poelzing and Rosenbaum, 2004) with equal-proportion segments of Endo, M and Epi cells (Yan et al., 1998). In the figures, time ran horizontally, space ran vertically. (E,F) APD distribution in both CTL and HF conditions for 1D strand models.

FIGURE 5. (A) Discontinuous CV at the PVJ in both CTL and HF conditions. The electrical excitation wave was conducted from the free-running end of the PF (starting from 0 mm), through the PVJ (18 mm) to the ventricular wall. (B) SF computed from the 1D strand model in both CTL and HF conditions. (C,D) VW computed from the 1D strand in the CTL and HF conditions. (E) VW recorded at the PVJ in both CTL and HF conditions.

FIGURE 6. (A) Left: Illustration of 3D wedge geometry of PF-ventricle model with d PF = 2.1 mm. Right: 3D whole canine ventricle model with fiber orientation, from which a 3D slab (marked in red box) was extracted as the 3D wedge model used in this study. (B) Electrical wave propagation from the PF, through PVJ to the ventricle in the CTL and HF conditions. (C) Time delay (△t) of the conduction through the PVJ in the CTL and HF conditions. △t was recorded from the PF to ventricle approximately 2 mm across the PVJ (Kanter et al., 1993). (D–F) Relationships between the CV (D), OS (E), SF (F) and the width of the PF strand at the PVJ.

FIGURE 7. Conduction failure at the PVJ with the PF width of 1.4 mm in the HF condition. (A) Model geometry with the selected points indicated as dots. Point I was at the middle of the PF. Point IV was at the middle of the Endo layer. Point V was at the middle of M layer. Point VI was at the middle of Epi layer. Points II and III represented the PF and Endo cell within the PVJ. (B–F) APs recorded from selected sites. (H) Demonstration of wave propagation failure at PVJ with the PF width of 1.4 mm in the HF condition.

FIGURE 8. (A–C) Excitation wave propagation in the CTL condition. APs at the middle of the thin (B) and thick (C) PF regions were recorded. (D–F) Excitation wave propagation in the HF condition. APs at the middle of the thin (E) and thick (F) PF regions were recorded.