Sodium-calcium exchange is essential for effective triggering of calcium release in mouse heart

Abstract

In cardiac myocytes, excitation-contraction coupling depends upon sarcoplasmic reticular Ca2+ release triggered by Ca2+ influx through L-type Ca2+ channels. Although Na+-Ca2+ exchange (NCX) is essential for Ca2+ extrusion, its participation in the trigger process of excitation-contraction coupling is controversial. To investigate the role of NCX in triggering, we examined Ca2+ sparks in ventricular cardiomyocytes isolated from wild-type (WT) and cardiac-specific NCX knockout (KO) mice. Myocytes from young NCX KO mice are known to exhibit normal resting cytosolic Ca2+ and normal Ca2+ transients despite reduced L-type Ca2+ current. We loaded myocytes with fluo-3 to image Ca2+ sparks using confocal microscopy in line-scan mode. The frequency of spontaneous Ca2+ sparks was reduced in KO myocytes compared with WT. However, spark amplitude and width were increased in KO mice. Permeabilizing the myocytes with saponin eliminated differences between spontaneous sparks in WT and KO mice. These results suggest that sarcolemmal processes are responsible for the reduced spark frequency and increased spark width and amplitude in KO mice. When myocytes were loaded with 1 mM fluo-3 and 3 mM EGTA via the patch pipette to buffer diadic cleft Ca2+, the number of sparks triggered by action potentials was reduced by 60% in KO cells compared to WT cells, despite similar SR Ca2+ content in both cell types. When EGTA was omitted from the pipette solution, the number of sparks triggered in KO and WT myocytes was similar. Although the number of sparks was restored in KO cells, Ca2+ release was asynchronous. These results suggest that high subsarcolemmal Ca2+ is required to ensure synchronous triggering with short spark latency in the absence of NCX. In WT mice, high subsarcolemmal Ca2+ is not required for synchronous triggering, because NCX is capable of priming the diadic cleft with sufficient Ca2+ for normal triggering, even when subsarcolemmal Ca(2+) is lowered by EGTA. Thus, reducing subsarcolemmal Ca2+ with EGTA in NCX KO mice reveals the dependence of Ca2+ release on NCX.

Figure 1

Ca²⁺ sparks in resting WT and NCX KO myocytes. (A) Confocal line-scan images (2 ms/line) recorded in quiescent WT (left) and KO (right) mouse ventricular myocytes loaded with fluo-3 AM. Note the decreased number of sparks in the KO image. (B) 3D plots on a spatiotemporal scale showing representative Ca²⁺ sparks recorded in WT (left) and KO (right) cells. (C) Mean spark amplitude (ΔF/F), FWHM, and frequency observed in quiescent WT (white bars, n = 10 cells from three WT mice) and NCX KO (black bars, n = 8 cells from four KO mice) myocytes. ^∗P < 0.05.

Figure 2

Ca²⁺ sparks in permeabilized WT and NCX KO myocytes. Confocal line-scan images (2 ms/line) of Ca²⁺ sparks recorded in representative WT (upper left) and KO myocytes (upper right) permeabilized with saponin and loaded with fluo-3 (resting free Ca²⁺ set to 60 nM). The bar graphs below show no difference in peak fluorescence (F/F₀), size (FWHM), or frequency of sparks (mean + SE) in seven cells from two WT mice, and seven cells from two KO mice. We obtained similar results in 25 WT and KO cells from three WT and three KO mice in which the resting Ca²⁺ was set to 100 nM (data not shown).

Figure 3

Ca²⁺ sparks evoked by APs in WT and NCX KO myocytes in the presence of 3 mM EGTA. (A) Representative APs stimulated by current commands and (B) corresponding rapid (0.24 ms/line) line-scan images recorded simultaneously in representative WT and NCX KO myocytes. Cells were loaded with 1 mM fluo-3 and 3 mM EGTA via the patch pipette. APs and images are also shown on a higher-resolution temporal scale (scale bar, 10 ms). Dashed line indicates the time when the earliest Ca²⁺ spark was activated. Arrowheads mark the positions where couplons failed to activate. Fluorescence intensities are reported in self-ratioed ΔF/F magnitude as indicated in the adjoining palette. (C) Ca²⁺ spark latency histograms (15-ms bins) constructed from line-scan images recorded in WT (left, n = 8 cells from four mice) and NCX KO (right, n = 8 cells from four mice) myocytes.

Figure 4

Ca²⁺ sparks evoked by APs in WT and NCX KO ventricular myocytes in the absence of EGTA. Representative APs stimulated by current commands (A) and corresponding rapid (0.24 ms/line) line-scan images (B) recorded simultaneously in representative WT and NCX KO myocytes loaded with 1 mM fluo-3 (but no EGTA) via the patch pipette. APs and images are also shown on a higher-resolution temporal scale (scale bar, 10 ms). Dashed line indicates the time when the earliest Ca²⁺ spark was activated. Arrowheads mark the positions on the line-scan images where couplons failed to activate. Fluorescence intensities are reported in self-ratioed ΔF/F magnitude as illustrated in the adjoining palette. (C) Ca²⁺ spark latency histograms (15-ms bins) constructed from line-scan images recorded in WT (left, n = 8 cells from four mice) and NCX KO (right, n = 8 cells from four mice) myocytes.

Figure 5

Latency of Ca²⁺ sparks evoked by APs in WT and KO myocytes Bar plots showing the mean latency of Ca²⁺ sparks identified during the 230-ms recording period. Confocal line scans (at 0.24 ms/line) were obtained as shown in Figs. 3 and 4 in WT and NCX KO myocytes. (A) shows averaged latencies for WT and KO myocytes buffered with 3 and 0 mM EGTA (via the patch pipette), respectively. (B) Latencies of sparks during the first 15 ms after the stimulus (early pool) obtained for the same conditions indicated in A. Regardless of the EGTA concentration, there was a significant increase in average latency in the NCX KO compared with WT mice (^∗∗P < 0.001 for KO versus WT in 0 mM EGTA and ^∗∗∗P < 0.0001 for KO versus WT in 3 mM EGTA), but only a trend toward increased latency of either cell type alone in 3 mM EGTA vs. 0 mM EGTA.

Figure 6

Spatially averaged Ca²⁺ transients in WT and NCX KO myocytes. (A and C) Superimposed AP-evoked Ca²⁺ transients obtained from representative WT and KO myocytes loaded with 1 mM fluo-3 and 3 mM EGTA (A) or 1 mM fluo-3 without added EGTA (C). (Insets) The same Ca²⁺ transients normalized to peak to illustrate the decay. (B and D) Bar plots show the peak ΔF/F of the spatially averaged Ca²⁺ transients recorded in WT (N = 8) and KO (N = 8) myocytes loaded with 1 mM fluo-3 and 3 mM EGTA (B) or 1 mM fluo-3 alone (D). ^∗∗P < 0.001 for KO versus WT.