Beat-to-beat oscillations of mitochondrial [Ca2+] in cardiac cells

Abstract

The Ca2+-sensitive photoprotein aequorin and the new green fluorescent protein-based fluorescent Ca2+ indicators 'ratiometric-pericam' were selectively expressed in the mitochondria, cytosol and/or nucleus of spontaneously beating ventricular myocytes from neonatal rats. This combined strategy reveals that mitochondrial [Ca2+] oscillates rapidly and in synchrony with cytosolic and nuclear [Ca2+]. The Ca2+ oscillations were reduced in frequency and/or amplitude by verapamil and carbachol and were enhanced by isoproterenol and elevation of extracellular [Ca2+]. An increased frequency and/or amplitude of cytosolic Ca2+ spikes was rapidly mirrored by similar changes in mitochondrial Ca2+ spikes and more slowly by elevations of the interspike Ca2+ levels. The present data unequivocally demonstrate that in cardiac cells mitochondrial [Ca2+] oscillates synchronously with cytosolic [Ca2+] and that mitochondrial Ca2+ handling rapidly adapts to inotropic or chronotropic inputs.

Fig. 1. Subcellular distribution of mitochondria-targeted aequorin. (A) Image of a cardiomyocyte expressing mit-GFP obtained by confocal microscopy (Rizzuto et al., 1996). The subcellular distribution of mit-GFP is indistinguishable from that of mit-Aeq. The profile of the fluorescent intensity along the line depicted in (A) is presented in (B); a and b are regions corresponding to the mitochondria as presented in (A). The level of fluorescence in regions of the cytoplasm devoid of mitochondria and in the nuclear region was indistinguishable from the level of fluorescence in non-transfected cells and was at least 50 (or more)-fold lower than in regions rich in mitochondria. The kinetics of [Ca²⁺]_c (C) and [Ca²⁺]_m (D) changes following stimulation with 1 µM angiotensin II are presented, in the presence (dotted trace) and absence (continuous trace) of 4 µM FCCP. In this experiment the cells were plated at low density to avoid the formation of a syncytium, the signal was integrated for 1 s and calibrated in terms of [Ca²⁺] as described by Brini et al. (1995). Under these conditions the oscillatory pattern of [Ca²⁺] is not evident.

Fig. 2. Reconstitution efficiency of targeted aequorins. HeLa cells and cardiomyocytes were transfected with chimeras that encode aequorin targeted either to the cytosol (A) or to mitochondria (B). Reconstitution of the functional aequorin was carried out, as described in detail in Materials and methods, by incubating the cells with coelenterazine in saline medium containing either CaCl₂ or EGTA. After 1 h of incubation, the total light output was measured by lysing the cells with digitonin, in the presence of 10 mM CaCl₂. *p <0.05. Values are mean ± SD (n = 8).

Fig. 3. Effect of extracellular [Ca²⁺] on cytosolic Ca²⁺ oscillations. Cells were transfected with cyt-Aeq. Reconstitution with coelenterazine was carried out in Ca²⁺-free medium. Where indicated, 1, 2 or 4 mM CaCl₂ was added to the perfusion medium. On the ordinate, the aequorin luminescence (counts/50 ms) is presented.

Fig. 4. Effect of extracellular [Ca²⁺] on mitochondrial Ca²⁺ oscillations. Cells were transfected with mit-Aeq. Other conditions are as in Figure 3.

Fig. 5. Effect of verapamil on cytosolic (A) and mitochondrial (B) Ca²⁺ oscillations. The perfusion buffer contained 2 mM CaCl₂. Where indicated, 20 µM verapamil was added.

Fig. 6. Fast kinetic analysis of cytosolic (A) and mitochondrial (B) Ca²⁺ oscillations. The data are details of the kinetics shown in Figure 4 and are presented on an expanded time scale.

Fig. 7. Effect of isoproterenol on cytosolic (A and B) and mitochondrial (C and D) Ca²⁺ oscillations. In (B) and (D) part of the kinetics of (A) and (C), respectively, are presented on an expanded time scale. Where indicated, 1 µM isoproterenol was added. The medium contained 2 mM CaCl₂.

Fig. 8. Average frequency of cytosolic and mitochondrial Ca²⁺ spikes before (white column) and after (black column) isoproterenol (1 µM) addition. Values are mean ± S.D. (n = 8). *p <0.05.

Fig. 9. Effect of 500 nM carbachol on cytosolic (A) and mitochondrial (B) Ca²⁺ oscillations. The medium contained 2 mM CaCl₂. Other conditions are as in Figures 3 and 4.

Fig. 10. Spontaneous [Ca²⁺] oscillations measured with ‘ratiometric-pericams’ targeted to the mitochondria. Typical image of a neonatal cardiac cell expressing the ratiometric-pericam targeted to the mitochondria (A). The cell was illuminated at 415 nm for 200 ms. (B) The cells were illuminated in sequence for 50 and 20 ms at 415 and 490 nm. The emitted light (at 510 nm) was collected and the ratio between the intensities of the two images (415/490) was calculated off line. On the left side the normalized fluorescence emission ratio is presented. Normalization was carried out by setting the minimum value to 0 and the maximum peak of the 415/490 ratio to 1.

Fig. 11. Spontaneous [Ca²⁺] oscillations measured with ‘ratiometric-pericams’ targeted to the nucleus and the mitochondria. (A) Typical image of neonatal cardiac cells expressing both the nuclear and mitochondrial ratiometric-pericam. The cells were illuminated at 415 nm for 200 ms. (B) Time course of the 415/490 (normalized) ratio changes in both compartments. The kinetics of a region of interest taken over mitochondria or nucleus are presented as continuous or dotted lines, respectively (B). Other conditions as in Figure 10B.