Cardiac cellular coupling and the spread of early instabilities in intracellular Ca2+

Abstract

Recent experimental and modeling studies demonstrate the fine spatial scale, complex nature, and independent contribution of Ca(2+) dynamics as a proarrhythmic factor in the heart. The mechanism of progression of cell-level Ca(2+) instabilities, known as alternans, to tissue-level arrhythmias is not well understood. Because gap junction coupling dictates cardiac syncytial properties, we set out to elucidate its role in the spatiotemporal evolution of Ca(2+) instabilities. We experimentally perturbed cellular coupling in cardiac syncytium in vitro. Coupling was quantified by fluorescence recovery after photobleaching, and related to function, including subtle fine-scale Ca(2+) alternans, captured by optical mapping. Conduction velocity and threshold for alternans monotonically increased with coupling. Lower coupling enhanced Ca(2+) alternans amplitude, but the spatial spread of early (<2 Hz) alternation was the greatest under intermediate (not low) coupling. This nonmonotonic relationship was closely matched by the percent of samples exhibiting large-scale alternans at higher pacing rates. Computer modeling corroborated these experimental findings for strong but not weak electromechanical (voltage-Ca(2+)) coupling, and offered mechanistic insight. In conclusion, using experimental and modeling approaches, we reveal a general mechanism for the spatial spread of subtle cellular Ca(2+) alternans that relies on a combination of gap-junctional and voltage-Ca(2+) coupling.

Figure 1

Perturbation and quantification of cell-cell coupling. The effects of gap junction agonist (4PB) and gap junction antagonist (heptanol) on cellular connectivity were quantified by FRAP. (A) Representative fluorescence images at different time points during the FRAP: before bleaching, at t = 0 s and at t = 20 s. Bleached cells are indicated by yellow arrows. Inset shows the applied FRAP protocol. (B) Averaged perturbation-relaxation recovery curves of 4PB (red), control (blue), and 0.25 mM heptanol-treated samples. Error bars are 95% confidence interval. (C) The time constant calculated from the recovery curve fitting for 4PB, control, and heptanol samples. Samples treated with 4PB and 0.5 mM heptanol have significantly different time constants compared to control samples. Error bars are 95% confidence interval; number of examined samples is shown for each group. Diffusion is inversely proportional to the calculated time constants.

Figure 2

Effects of coupling on propagation. (A) CV measured at 30°C, 1 Hz pacing for each group in the rectangular samples; last bar shows recovery (wash out) after heptanol treatment. Error bars indicate standard deviation; (^∗) significance at p < 0.05; the 95% CI for the slope of a regression line was (23.7, 46.9). (B) Example activation maps for the four groups; color indicates time of activation (0 to 0.1 s), isochrones (in black) are 0.02 s apart. Ca²⁺ transients are shown from each sample at the ^∗ location. (C) Estimated relative effects of heptanol on CV via suppression of gNa and via gap junction block. The two diamonds indicate actual measured CV values (as % of control, from panel A) for the two heptanol concentrations used here; the three curves show expected reduction in CV via gNa block alone: analytical estimate (black), numerical solution (red), and our experimental data with TTX block (blue). See text and Fig. S1 for further details. (D and E) No significant effects of 4PB and low doses of heptanol on Ca²⁺ transients: CTD80 and rise time (mean ± SE) are not significantly different between the groups (p = 0.34 for 4PB effects and p = 0.16 for heptanol effects on CTD80); power analysis with bootstrapping would have uncovered significant changes if CTD80 were altered by >18.7%.

Figure 3

Rate-dependent evolution of fine-scale Ca²⁺ alternans. (A) Resolving fine-scale alternans while imaging a large FOV: (top) space-time plot of persistent (over 55 beats) fine-scale alternation at the cellular level at 4 Hz pacing, detected while imaging FOV = 2.2 × 2.2 cm²; color indicates Ca²⁺ AR (%), multiplied by representative phase, (−1, +1); (bottom) plot of fine-scale spatially discordant alternans along a line from the same sample. (B–D) Representative spatial maps of fine-scale Ca²⁺ alternans at steady state under different pacing frequencies and cellular coupling conditions: 0.5 mM heptanol (B), control (C), and 4PB samples (D). Pacing electrode is on the right side. Color indicates AR (%) at different spatial locations. Red and blue identify opposite phase; dark green areas are regions with no detected alternans; white areas identify conduction blocks. Traces on the right are from the location identified by a (^∗) at pacing frequencies 2.8, 3.4, and 4.0 Hz.

Figure 4

Linking dynamic properties of Ca²⁺ alternans to FRAP-quantified diffusion. (A) CV as a function of the FRAP-measured diffusion factor (1/τ, 1/s) for the four experimental groups. (B) Breakpoint frequency (frequency of failure to follow 1:1) as a function of the diffusion factor (1/τ, s). (C) Evolution of the areas exhibiting local Ca²⁺ alternans as a function of pacing frequency: green for low coupling (0.5 mM heptanol), black for intermediate coupling (control), and blue for high coupling (4PB); arrows indicate the average breakpoint frequency; error bars indicate standard error. (D) Space-averaged AR for early fine Ca²⁺ alternans (<2 Hz) as a function of coupling/diffusion. (E) Area of fine Ca²⁺ alternation at low frequencies (<2 Hz) as a function of coupling. (F) Proportion of samples (%) that exhibited large-scale 2:2 alternans upon breakpoint frequency (the remaining samples went directly into 2:1 block). Sample numbers for A∼F are n_0.5 mM = 17, n_0.25 mM = 14, n_control = 28, n_4PB = 18; data are mean ± 95% confidence interval.

Figure 5

Theoretical explanation of the effects of coupling on the spatial properties of Ca²⁺ alternans. Using a computational model of cardiac tissue, an island of Ca²⁺ alternans (inset) was simulated under different coupling conditions. Quantification of spatial patterns by imposing a threshold (A) average AR under 4 Hz pacing as a function of diffusion properties (B); area of Ca²⁺ alternation above a threshold as a function of diffusion properties (C). Color (blue to red) indicates low to high diffusion. Compare panel B to Fig. 4D, and panel C to Fig. 4, E and F. (D). Schematic illustration of the effective diffusion of Ca²⁺ alternans via D_v: in a cell pair or a cable. (E) In the cell/region exhibiting instability (gray), Ca²⁺ alternans drive V_m alternans, which propagate according to D_v. In the distant cell/region, V_m alternans cause Ca²⁺ alternans, thus effectively mediating diffusion of Ca²⁺ alternans. (F) Effects of V_m-Ca²⁺ coupling on the diffusion of Ca²⁺ alternans: same simulations as in A, but under low V_m-Ca²⁺. Ca²⁺ alternans do not spread outside the unstable region (E) because of negligible V_m alternans amplitude and low V_m-Ca²⁺ coupling in this case (G).