In cardiomyocytes, excitationÐcontraction (EC) coupling is accompanied by a transient rise in the intracellular Ca²⁺ concentration, that is to a varying degree brought about by Ca²⁺ release from the sarcoplasmic retiuculum (SR), which is mediated by the opening of ryanodine receptors (RyR) in the SR membrane. The global Ca²⁺ transient is built-up by the synchronised recruitment of local Ca²⁺ release events, the so-called Ca²⁺ sparks, that are generated by gating of clusters of RyR (up to 100 channels). Activation of these Ca²⁺ sparks during EC coupling is a crucial step in the activation of myocyte contraction and depends on the intimate interaction between L-type Ca²⁺ channels in the plasma membrane on one hand and RyR in the membrane of the SR as the counterpart. Under various pathological conditions this interaction has been hypothesised to be partially disrupted. For example, in cardiomyocytes from spontaneously hypertensive rats, detailed analysis of Ca²⁺ sparks revealed that these events had a higher amplitude than in the control normotensive rats (Shorofsky et al. 1999).

We analysed Ca²⁺ sparks from atrial cardiomyocytes isolated from healthy human tissue obtained during by-pass operations in order to characterise these elementary Ca²⁺ events in greater detail. After enzymatic digestion of a small piece of human tissue, isolated cells were loaded either with fluo 3-AM or fluo 4-AM and the spatio-temporal properties of Ca²⁺ sparks were recorded with a Noran confocal microscope using linescan acquisition mode (278 lines per seconds). Sequences of linescan images were analysed using NIH Image and Igor software. In the majority of cells studied, Ca²⁺ sparks had a time to peak of 35.4 ± 2.1 ms and a mean amplitude of 132 ± 13 nM (all data presented as means ± S.E.M.). The duration, T₅₀, was measured at half of the Ca²⁺ spark amplitude and was equal to 37 ± 2.6 ms and their spatial spread (full width at half-maximum) was 4.2 ± 0.15 mm (n = 44 sparks in n = 5 cells). However, it was found that larger subcellular events also co-existed with ‘normal’ Ca²⁺ sparks. These events had a time to peak of 43.2 ± 4 ms and a mean amplitude of 118.6 ± 9 nM (n = 19 sparks in 5 cells). T₅₀ was 33.7 ± 4.3 ms and their spatial spread increased to 9.75 ± 0.9 mm and was significantly different from the ‘normal’ Ca²⁺ sparks (P ²le³ 0.05, unpaired t test). We were surprised to find such a population of ‘big sparks’ in myocytes from ‘healthy’ (i.e. non-hypertrophied) human hearts, since large amplitude Ca²⁺ sparks have recently been reported to only occur under pathological conditions (Shorofsky et al. 1999).

Results such as those presented here will establish a baseline for the properties of Ca²⁺ sparks in human atrial tissue and will allow for the detailed comparison of their properties with the characteristics of Ca²⁺ sparks found in human hearts obtained from patients suffering various heart diseases.

All procedures accord with current local guidelines.