Methods Cardiomyocytes were studied isolated from rats with heart failure 4–6 weeks post in vivo AAV9SERCA (2×1011 drp) gene transfer (HF+S). Cells from untreated failing (HF) and non failing (NF) hearts served as controls. Spontaneous Ca²⁺ spark generation was measured in fluo4-loaded cells by scanning confocal microscopy during a quiescent 30 s period after baseline stimulation. Field stimulated calcium transients were measured at 0.5 Hz, followed by caffeine application to measure SR Ca²⁺ load. The ratio of sparks:SR load and Leak Index reflecting total SR Ca²⁺ leak (spark frequency × amplitude × width × duration) were calculated. Baseline and isoproterenol (ISO)-induced triggered activity were measured in parallel cell studies, and ryanodine receptor (RyR) phosphorylation was measured.

Results SERCA2a gene transfer normalised SR Ca²⁺ load of failing myocytes (HF+S 5.86±0.64 vs HF 4.24±0.32 p<0.05 (NF 5.34±0.38)) with recovery of Ca²⁺ transient relaxation kinetics (R50 (ms): HF+S 174±14, HF 292±42, NF 179±23 p<0.01). Spark frequency was unchanged between HF+S (2.9±0.4) and HF (2.5±0.4), with both significantly higher than NF (1.2±0.2 p<0.01). Interestingly spark amplitude was significantly reduced in HF+S compared with HF, with no change in spark decay kinetics, leading to a significant reduction in total SR leak. Leak index: NF234±76, HF 1630±358, HF+S 683±192 (p<0.05). Total RyR was reduced by 20% in failing hearts, with a significant increase in the proportion of RyR-P2809 and RyR-P2815 compared to non-failing controls (p<0.05). These RyR-P and total RyR changes were reversed to normal after SERCA2a gene transfer to failing hearts. Despite increased SR content and unaltered spark frequency, myocytes from HF+S hearts were more stable in vitro, with significantly fewer basal and ISO-induced aftercontractions compared to HF (p<0.05).

Conclusion SERCA2a gene therapy reduced total SR Ca²⁺ leak of failing myocytes whilst concurrently increasing SR Ca²⁺ load to normal levels. Leak reduction was dependent upon altered spark characteristics as spontaneous spark frequency was unchanged, and reduced RyR phosphorylation supports resetting of SR Ca²⁺ leak threshold after SERCA2a gene transfer. SERCA2a gene therapy reduced ISO-induced triggered activity in vitro and ISO-induced arrhythmias in vivo, and may represent a novel antiarrhythmic strategy in heart failure.