The pacemaker cells of the heart initiate the heartbeat, sustain the circulation, and dictate the rate and rhythm of cardiac contraction. Circulatory collapse ensues when these specialized cells are damaged by disease, a situation that currently necessitates the implantation of an electronic pacemaker. Here we report the use of viral gene transfer to convert quiescent heart-muscle cells into pacemaker cells, and the successful generation of spontaneous, rhythmic electrical activity in the ventricle in vivo. Our results indicate that genetically engineered pacemakers could be developed as a possible alternative to implantable electronic devices.

Membrane current abnormalities have been described in human heart failure. To determine whether similar current changes are observed in a large animal model of heart failure, we studied dogs with pacing-induced cardiomyopathy. Myocytes isolated from the midmyocardium of 13 dogs with heart failure induced by 3 to 4 weeks of rapid ventricular pacing and from 16 nonpaced control dogs did not differ in cell surface area or resting membrane potential. Nevertheless, action potential duration (APD) was significantly prolonged in myocytes isolated from failing ventricles (APD at 90% repolarization, 1097 +/- 73 milliseconds [failing hearts, n = 30] versus 842 +/- 56 milliseconds [control hearts, n = 25]; P < .05), and the prominent repolarizing notch in phase 1 was dramatically attenuated. Basal L-type Ca2+ current and whole-cell Na+ current did not differ in cells from failing and from control hearts, but significant differences in K+ currents were observed. The density of the inward rectifier K+ current (IKl) was reduced in cells from failing hearts at test potentials below -90 mV (at -150 mV, -19.1 +/- 2.2 pA/pF [failing hearts, n = 18] versus -32.2 +/- 5.1 pA/pF [control hearts, n = 15]; P < .05). The small outward current component of IKl was also reduced in cells from failing hearts (at -60 mV, 1.7 +/- 0.2 pA/pF [failing hearts] versus 2.5 +/- 0.2 pA/pF [control hearts]; P < .05). The peak of the Ca(2+)-independent transient outward current (Ito) was dramatically reduced in myocytes isolated from failing hearts compared with nonfailing control hearts (at +80 mV, 7.0 +/- 0.9 pA/pF [failing hearts, n = 20] versus 20.4 +/- 3.2 pA/pF [control hearts, n = 15]; P < .001), while the steady state component was unchanged. There were no significant differences in Ito kinetics or single-channel conductance. A reduction in the number of functional Ito channels was demonstrated by nonstationary fluctuation analysis (0.4 +/- 0.03 channels per square micrometer [failing hearts, n = 5] versus 1.2 +/- 0.1 channels per square micrometer [control hearts, n = 3]; P < .001). Pharmacological reduction of Ito by 4-aminopyridine in control myocytes decreased the notch amplitude and prolonged the APD. Current clamp-release experiments in which current was injected for 8 milliseconds to reproduce the notch sufficed to shorten the APD significantly in cells from failing hearts. These data support the hypothesis that downregulation of Ito in pacing-induced heart failure is at least partially responsible for the action potential prolongation. Because the repolarization abnormalities mimic those in cells isolated from failing human ventricular myocardium, canine pacing-induced cardiomyopathy may provide insights into the development of repolarization abnormalities and the mechanisms of sudden death in patients with heart failure.