Ionic mechanisms of electrophysiological heterogeneity and conduction block in the infarct border zone

Abstract

The increased incidence of arrhythmia in the healing phase after infarction has been linked to remodeling in the epicardial border zone (EBZ). Ionic models of normal zone (NZ) and EBZ myocytes were incorporated into one-dimensional models of propagation to gain mechanistic insights into how ion channel remodeling affects action potential (AP) duration (APD) and refractoriness, vulnerability to conduction block, and conduction safety postinfarction. We found that EBZ tissue exhibited abnormal APD restitution. The remodeled Na(+) current (I(Na)) and L-type Ca(2+) current (I(Ca,L)) promoted increased effective refractory period and prolonged APD at a short diastolic interval. While postrepolarization refractoriness due to remodeled EBZ I(Na) was the primary determinant of the vulnerable window for conduction block at the NZ-to-EBZ transition in response to premature S2 stimuli, altered EBZ restitution also promoted APD dispersion and increased the vulnerable window at fast S1 pacing rates. Abnormal EBZ APD restitution and refractoriness also led to abnormal periodic conduction block patterns for a range of fast S1 pacing rates. In addition, we found that I(Na) remodeling decreased conduction safety in the EBZ but that inward rectifier K(+) current remodeling partially offset this decrease. EBZ conduction was characterized by a weakened AP upstroke and short intercellular delays, which prevented I(Ca,L) and transient outward K(+) current remodeling from playing a role in EBZ conduction in uncoupled tissue. Simulations of a skeletal muscle Na(+) channel SkM1-I(Na) injection into the EBZ suggested that this recently proposed antiarrhythmic therapy has several desirable effects, including normalization of EBZ effective refractory period and APD restitution, elimination of vulnerability to conduction block, and normalization of conduction in tissue with reduced intercellular coupling.

Fig. 1.

Schematic of the canine epicardial model, with the remodeled epicardial border zone (EBZ) currents in shaded circles. CT_KCl, K⁺-Cl⁻ cotransporter; CT_NaCl, Na⁺-Cl⁻ cotransporter; I_NaL, late Na⁺ current; I_Na, fast Na⁺ current; I_Nab, background Na⁺ current; I_Na,Ca,i, Na⁺/Ca²⁺ exchanger current (localized to myoplasm); I_Cab, background Ca²⁺ current; I_pca, sarcolemmal Ca²⁺ pump; I_to1, 4-aminopyridine-sensitive transient outward K⁺ current; I_Kr, fast delayed rectifier K⁺ current; I_Ks, slow delayed rectifier K⁺ current; I_K1, time-independent K⁺ current; I_Kp, plateau K⁺ current; I_NaK, Na⁺-K⁺ pump current; I_Ca,L, L-type Ca²⁺ current; I_to2, Ca²⁺-dependent transient outward Cl⁻ current; I_Na,Ca,ss, Na⁺/Ca²⁺exchanger current [localized to the sarcoplasmic reticulum (SR) subspace (SS)]; CaMKII, Ca²⁺/calmodulin-dependent kinase II; PLB, phospholamban; SERCA, sarco(endo)plasmic reticulum Ca²⁺-ATPase; CSQN, calsequestrin; I_rel, Ca²⁺ release from the junctional SR (JSR); SS(CaL), I_CaL SS; I_diff,ss, ion diffusion [from SS to local SS(CaL)]; I_tr, Ca²⁺ transfer [network SR (NSR) to JSR]; I_leak, NSR leak; I_diff, Ca²⁺ diffusion [SS(SR) to myoplasm]. Shaded arrows indicate CaMKII targets. Additional model details can be found in Refs. and , the online Supplemental Material, and http://rudylab.wustl.edu/.

Fig. 2.

Comparison of model and experimental data. A: normal zone (NZ) and EBZ maximum action potential (AP) upstroke velocity (dV/dt_max) and time constant of recovery of AP dV/dt_max (τ_recovery) in model and experimental data (26). dV/dt_max properties were determined as described in Ref. . Membrane voltage (V_m) was held at 0 mV for 500 ms, followed by varying time intervals at the holding potential (V_hold). APs were elicited by the application of a stimulus after the release of voltage clamp at V_hold. Fully recovered dV/dt_max was determined for an AP elicited after 500 ms at V_hold = −110 mV. τ_recovery was determined by an exponential fit to the time course of recovery of AP dV/dt_max at V_hold = −80 mV. B: I_Ca,L-voltage [current-voltage (I-V)] relationship and time constant of voltage-dependent inactivation (τ_VDI) in model and experimental data (1). C and D: NZ and EBZ I-V relationship of I_to1 (C) (26) and I_to2 (D) (2) in model and experimental data. E: ratio (EBZ to NZ) of I_Ks density (21), I_K1 density (26), I_Kr density (21), and I_Kr (21) time constant of activation (τ_activation) in model and experimental data.

Fig. 3.

A: model NZ and EBZ cell APs [S1 cycle length (CL_S1) = 1 s and extracellular K⁺ concentration ([K⁺]_o) = 4 mM]. B: NZ and EBZ AP duration (APD; CL_S1 = 1 s and [K⁺]_o = 4 mM) at 50% and full (98%) repolarization (APD₅₀ and APD₉₈) in model and experimental data (20, 26, 39). C: model NZ ([K⁺]_o = 4.5 mM) and EBZ ([K⁺]_o = 7.6 mM) APs (CL_S1 = 1 s) for the central cell in homogeneous strand simulations.

Fig. 4.

A: APD restitution measured in the central cell of homogenous NZ (NZ_HOMO) and homogenous EBZ (EBZ_HOMO) strands (CL_S1 = 1 s). DI, diastolic interval. B–E: V_m (B), I_Na (C), I_Ca,L (D), and I_Kr (E) as a function of DI in NZ (left) and EBZ (right) strands. APs were time shifted to align dV/dt_max for each DI. APD was calculated at 95% repolarization. PRR, postrepolarization refractoriness.

Fig. 5.

A: rate dependence in heterogenous NZ-EBZ (NZ-EBZ_HET) strand simulations. Strands were composed of 48 NZ and 48 EBZ cells, with stimulation at the NZ end. APD is shown for central NZ (cell 24) and central EBZ (cell 72) cells. B: vulnerable window (VW) as a function of S2 coupling interval (CI_S2) for a range of CL_S1. C and D: V_m, I_Na, and I_Ca,L from NZ (cell 24) and EBZ (cell 72) cells CL_S1 = 300 ms (C) and 1,000 ms (D). APD was calculated at 95% repolarization.

Fig. 6.

A: VW in NZ-EBZ_HET strands during pacing at fast CL_S1. Strands were composed of 48 NZ and 48 EBZ cells. Stimuli were applied at the NZ end. B–D: NZ (cell 24) and EBZ (cell 72) V_m, I_Na, and I_Ca,L in heterogeneous strands during pacing at fast CL_S1 showing representative patterns of conduction and block. B: 3:2 EBZ conduction block (CL_S1 = 270 ms). C: 2:1 EBZ conduction block (CL_S1 = 250 ms). D: 2:1 NZ and EBZ conduction block (CL_S1 = 200 ms). APD was calculated at 95% repolarization.

Fig. 7.

Role of I_K1 in NZ and EBZ conduction. A and B: safety factor (SF; A) and conduction velocity (CV; B) as a function of gap junction conductance (g_j) in NZ_HOMO, EBZ_HOMO, and EBZ_HOMO strands with normal I_K1 restored. C–E: V_m (C), I_K1 (D), and I_Na (E) for propagating APs in EBZ_HOMO strands with remodeled and normal I_K1 (greatly reduced gap junction coupling, g_j = 0.069 μS). The SF, CV, and APs for each strand type are shown at the central cell (cell 48).

Fig. 8.

Role of I_Ca,L and I_to1 in NZ and EBZ conduction. A: SF for propagation as a function of gap junction coupling in NZ_HOMO (left) and EBZ_HOMO (right) strands. B: V_m of upstream and downstream cells during propagation in NZ_HOMO (left) and EBZ_HOMO (right) strands at the minimum g_j (critical g_j) for which propagation occurred. C: activation of I_Na, I_to1, and I_Ca,L for a propagating AP at the critical g_j in NZ_HOMO (left) and EBZ_HOMO (right) strands. Normal I_to1 was included in the EBZ_HOMO strand simulation to assess its potential role in propagation. D: charge contribution (Q) of I_Na, I_Ca,L, and I_to1 to propagation at the critical g_j in NZ_HOMO (left) and EBZ_HOMO strands with normal I_to1 restored (right). SF, CV, and APs for each strand type are shown at the central cell (cell 48).

Fig. 9.

A–C: effects of skeletal muscle Na⁺ channel (SkM1-I_Na) addition on steady-state I_Na availability (A), APD restitution (B) and SF (C) in NZ_HOMO, EBZ_HOMO, and EBZ_HOMO + SkM1-I_Na strands. D: effect of SkM1-I_Na addition on VW during S1-S2 protocols for CL_S1 = 300, 500, and 1,000 ms in NZ-EBZ_HET strands after SkM1-I_Na addition. E: VW at fast CL_S1 in NZ-EBZ_HET strands after SkM1-I_Na addition to the EBZ region.