Arrhythmogenic late Ca2+ sparks in failing heart cells and their control by action potential configuration

Significance Sudden cardiac death in heart failure is a major unsolved clinical problem that is linked to the development of a spontaneous arrhythmia. Early afterdepolarizations (EADs) are an arrhythmogenic mechanism, but the cellular trigger for EADs in heart failure is unclear. We show that the reduction in synchronous Ca2+ release early in the action potential (AP) of failing cardiac myocytes promotes the appearance of late Ca2+ sparks which can propagate, forming Ca2+ ripples and waves. These, in turn, produce an inward sodium–calcium exchange current which opposes AP repolarization. Restoration of AP phase 1 repolarization improved Ca2+ release synchrony and reduced late Ca2+ spark rate, suggesting a different approach to reducing the risk of sudden death in heart failure.

Action Potential Clamp Recording. Action potential (AP) clamp involves applying a selected AP waveform as the voltage command to a cell under voltage clamp. The AP waveform may be recorded from the same cell (in current clamp), a typical AP from the same species, or even modified to probe the role of selected ionic currents. During the applied AP, LTCC open in response to the AP waveform as if it were produced endogenously and therefore the Ca 2+ transient that is produced reveals the functional consequences of the AP waveform on excitation-contraction coupling and Ca 2+ release. For example, changing the AP waveform to mimic reduced repolarization rate (2), frequency dependent AP duration shortening (3), and heart failure(4), have all been used previously to reveal the role of AP changes on the Ca 2+ transient and contraction.
Current Clamp Vm Recording. Cells were electrically stimulated by 2 ms depolarizing current injections at 1.3x threshold at 1 Hz. Threshold current was found by progressively increasing pulse amplitude in 0.2 nA increments until APs were consistently evoked.
Calcium imaging analysis. Non-cell background fluorescence from an area adjacent to the cell was subtracted from recordings. Variations in fluorescence due to dye loading was minimized by normalizing fluorescence (F) to resting fluorescence during a 50 ms quiescent period immediately before stimulation (F0). The F/F0 recording was converted into units of [Ca 2+ ] using the self-ratio method: Where K is the in vivo affinity of Fluo-4 for Ca 2+ (Kd ~1000 nmol/L), R is the self-ratio fluorescence (F/F0), and [Ca 2+ ]rest is the resting Ca 2+ concentration (~100 nmol/L) (5).
Low-frequency time-averaged fluorescence in xt line scan recordings was subtracted from the F/F0 recording using a low-pass quadratic Savitsky-Golay filter (window size ~101-151 ms) applied along the t dimension, for every point in the x dimension. These filter values reduced background fluorescence variation due to the underlying Ca 2+ transient. LCS were then detected using an optimal filter object detection algorithm implemented in MATLAB as described previously (6).    Figure S3. Changes in the Ca 2+ transient and AP duration in HF. A. As might be expected, HF myocyte Ca 2+ transients had a reduced rate of rise and decline compared to CON. B. Although there was a tendency to reduced Ca 2+ transient amplitude in HF, this was not significant at the group level. C The duration of the AP at 20% repolarization was increased in HF cells and D at 90% repolarization. *p<0.05; ** p<0.01, n/N= 16/6 CON ; 14/5 HF.

Figure S4
LCS and Ca 2+ ripples underlying the oscillatory Ca 2+ release shown in Fig. 2A. Although many LCS can be seen in the upper panels, image processing (lower panel) of the selected region shows numerous chevron-like patterns in underlying LCS activity. Such chevron shapes indicate the presence of Ca 2+ ripples (10) where propagation of CICR via cytoplasmic Ca 2+ recruits additional LCS sites. This is clarified by the added lines for some of the Ca 2+ ripples in the lower panel. Note also the very low amplitude chevron shapes that probably reflect Ca 2+ ripples below the confocal plane. Figure S5. Injection of an Inet balancing current halts AP repolarization. A A model generated A is shown by the black line. B Inet was calculated for the AP in A using the equation Inet= -Cm.dVm/dt (see text). Arrows and associated labels indicate the amplitude of Inet at the indicated points in the AP. C The black line shows the rate of change in membrane potential during the normal AP (dVm/dt). In subsequent simulations, a constant 'balancing' current of equal amplitude but opposite polarity to Inet was injected at the times indicated by arrows. This immediately halted repolarization, as shown by the reduction of dVm/dt to 0 V/s and the cessation of repolarization in the AP (shown by the red, green and blue traces in A). Figure S6. Ca 2+ dependence of the LCS evoked inward current in Fig. 2. Note the weak voltage dependence of the change in current but strong effect of Ca 2+ as expected for NCX (22). Blue data recorded at Vm = +20 mV; Black at Vm = 0 mV and red at Vm = -20 mV. Table S1. Organ weights of CON and HF rabbits. N = 5 CON and 13 HF rabbits. Unpaired t-test.