Modeling tissue- and mutation- specific electrophysiological effects in the long QT syndrome: role of the Purkinje fiber

Abstract

Congenital long QT syndrome is a heritable family of arrhythmias caused by mutations in 13 genes encoding ion channel complex proteins. Mounting evidence has implicated the Purkinje fiber network in the genesis of ventricular arrhythmias. In this study, we explore the hypothesis that long QT mutations can demonstrate different phenotypes depending on the tissue type of expression. Using computational models of the human ventricular myocyte and the Purkinje fiber cell, the biophysical alteration in channel function in LQT1, LQT2, LQT3, and LQT7 are modeled. We identified that the plateau potential was important in LQT1 and LQT2, in which mutation led to minimal action potential prolongation in Purkinje fiber cells. The phenotype of LQT3 mutation was dependent on the biophysical alteration induced as well as tissue type. The canonical ΔKPQ mutation causes severe action potential prolongation in both tissue types. For LQT3 mutation F1473C, characterized by shifted channel availability, a more severe phenotype was seen in Purkinje fiber cells with action potential prolongation and early afterdepolarizations. The LQT3 mutation S1904L demonstrated striking effects on action potential duration restitution and more severe action potential prolongation in Purkinje fiber cells at higher heart rates. Voltage clamp simulations highlight the mechanism of effect of these mutations in different tissue types, and impact of drug therapy is explored. We conclude that arrhythmia formation in long QT syndrome may depend not only on the basis of mutation and biophysical alteration, but also upon tissue of expression. The Purkinje fiber network may represent an important therapeutic target in the management of patients with heritable channelopathies.

Figure 1. Differential response of VM (panel A) and PF (panel B) to mutation in IKs and IKr channels, leading to LQT1 and LQT2, respectively.

Compared to wild type (WT, black trace), heterozygous LQT1 (blue trace) and LQT2 (red trace), modeled as a 50% reduction in whole cell conductance, VM APs are prolonged, while PF are mildly affected. Homozygous LQT1 (100% reduction in IKs conductance, dashed blue trace) leads to substantial APD prolongation in VM, and trivial further prolongation in PF.

Figure 2. Arrhythmias in Andersen-Tawil syndrome, modeled as dominant negative suppression of IK1 current, have different mechanisms in VM (panel A) versus PF (panel B) myocytes.

For VM, compared to WT (black trace) reduction of IK1 conductance by 90% (blue) leads to a large delayed afterdepolarization. Further reduction to 95% (red) increases the size of the DAD, triggering repetitive action potentials. For PF, reduction of IK1 conductance maximum diastolic potential is less polarized, leading to increased automaticity. Unstimulated PF APs show basic cycle length of automaticity of 2250 msec, which decreases to 1260 with 90% IK1 reduction.

Figure 3. VM and PFC carrying the SCN5A ΔKPQ mutation exhibit similarly severe APD prolongation.

Panel A and B) Effect of a 2 second pause for VM and PFC, respectively, compared with 1000 msec (last beat of drive train shown for WT, black trace, and mutant cells, blue trace). Panels C and D) Steady state rate adaptation for VM and PFC, respectively (30 BPM: blue trace, 40 BPM: red trace, 60 BPM: black trace). E) Summary results for APD prolongation for each cell type for different stimulus protocols; VM: black bars, PF: gray bars.

Figure 4. Differential effect on APD for VM versus PF of the SCN5A F1473C mutation.

Panels A and B) Effect of a 2 second pause for VM and PFC, respectively, compared with 1000 msec (last beat of drive train shown for WT, black trace, and mutant cells, blue trace). APD in the F1473C shows pause-dependent increase, with spontaneous oscillations in membrane potential from sequential early afterdepolarizations seen post-pause for PF but not VM. Panels C and D) Rate adaptation for VM and PFC, respectively (30 BPM: blue trace, 40 BPM: red trace, 60 BPM: black trace. In VM cells, substantial APD prolongation is seen (on the order of that seen with ΔKPQ mutation) with pronounced bradycardia dependence. In PF APs, severe APD prolongation is seen at 60 BPM, for 40 BPM stable alternans are seen (with a small early afterdepolarization on every other beat), and at 30 BPM the membrane fails to repolarize.

Figure 5. The S1904L mutation, characterized clinically by arrhythmia at low levels of exertion, simulated at clinically relevant pacing frequencies.

The mutation exhibits minimal response in VM (panel A, traces overlap) and greater APD prolongation in PF (panel B), with APD alternans seen at 120 BPM (lower row). WT cells: black trace, S1904L: blue trace.

Figure 6. Voltage clamp protocols demonstrate themechanism of late current generation in LQT3 mutations.

Panel A illustrates the voltage clamp waveform used. Panel Bdemonstrates the important role of resting membrane potential and plateaupotential on degree of late sodium current in ΔKPQ (red bar)and F1473C mutations (blue bar) compared to WT (black bar).For a “VM-like” voltage clamp protocol (holding −100 mV,step potential 20 mV, top row), similar persistent current isseen for KPQ and F1473C mutations. When holding potential ischanged to −80 mV mimicking PF AP resting potential, second row,persistent current is less for the KPQ mutation than for the F1473C mutation (whichhas increased availability at this potential). Third andfourth rows show the individual effect of altering plateau and resting potentials.Panel C) Response of step potentials simulating VM (black)and PF (blue) in cells carrying the S1904L mutation.The “PF-like” voltage clamp waveform produces more inwardcurrent at all time points, leading to more accumulated inward charge as demonstratedbelow in the bar graph.