The Normal Cardiac Pacemaker

The mammalian sinoatrial node was identified structurally in 1907 by Keith and Flack,⁵ and its function as the primary source of the cardiac impulse was demonstrated electrophysiologically by Lewis et al in 1910.⁶ These investigators recognized that they were dealing with a heterogeneous structure that appeared to serve as the dominant site of origin of the normal heartbeat. Yet the mechanisms determining sinoatrial node function remained a mystery for decades. One school of thought held that the node—and indeed the entire cardiac conduction system—consisted of specialized neural fibers, with a neural impulse initiating the heartbeat.⁷ That sinus rhythm persisted after cardiac denervation⁸ argued against central neural origin, although a role for local neuronal activity as the cause of the heartbeat was a continuing belief.⁷

The modern era of molecular biophysics has determined the why and the wherefore of the origin of cardiac impulses, although the origin of the sinoatrial node cells themselves remains a subject of active investigation.⁹ As is shown in Figure 26-1, A, multiple currents contribute to the sinoatrial node action potential.^10–13 That the resting membrane potentials of normal sinoatrial node cells are more depolarized than those of most other cardiac myocytes is the result of a small inward rectifier current, I_K1. Action potentials initiated at these voltages have calcium rather than sodium as the inward charge carrier. The result is a slowly propagating impulse similar to that in the normal AV node, the AV valves, and the coronary sinus.

As sinoatrial node cells repolarize to their maximum diastolic potentials, their hyperpolarization permits the opening of HCN (for Hyperpolarization-activated Cyclic Nucleotide gated) channels, through which an inward sodium current, referred to as I_f, flows and begins to depolarize the membrane (Figure 26-1, A, B).^10–12 A major contributor to pacemaker activity is the calcium clock mechanism, which depends on calcium cycling between intracellular uptake and release sites, as well as calcium entry via L- and T-type Ca channels.^13,14 This affects the activity of the Na/Ca exchanger, which directly modulates phase 4 depolarization.^13,14 All this activity occurs against the backdrop of outward (repolarizing) potassium current, such that any increase in inward current and/or decrease in outward current will increase the slope of phase 4 depolarization and the pacemaker rate.

The sympathetic and parasympathetic nervous systems are the principal modulators of phase 4 depolarization. Their respective neurohumors, norepinephrine and acetylcholine, act via β-adrenergic or muscarinic receptor-G protein-linked pathways to increase (β-adrenergic) or decrease (muscarinic) cyclic adenosine monophosphate (cAMP) synthesis (see Figure 26-1, B). Cyclic AMP binding to a site near the carboxy terminus of the HCN channel shifts channel activation positively, resulting in an increase in phase 4 depolarization.¹²

Whereas all the currents described contribute to phase 4 depolarization and to pacemaker rate, the initiator of the process appears to be I_f.10,11 That all pacemaker activity is not attributable to I_f has been demonstrated in experiments using I_f-blocking drugs. In both experimental models and the clinic, the result is significant slowing of sinus rate but not termination of the rhythm.^15,16 Indeed, the inability to suppress sinus node function using any one blocker highlights the redundancy within the system such that it continues to function even under challenge.

The HCN channels that initiate pacemaker function have 6 transmembrane-spanning domains, are ubiquitous in the heart, and are found elsewhere in the body as well (see Figure 26-1, B). Four channel isoforms, HCN1-4, are present, with 4 and 1 in the sinoatrial node, 2 in much of the conducting system and the myocardium, and 3 largely in neural tissues.¹² Each has differing activation and deactivation kinetics and differing cAMP sensitivity. The critical aspects of these channels with regard to pacemaker activity include their activation on hyperpolarization and their modulation by autonomic neurohumors.

To summarize, the primary biological pacemaker is structurally and functionally complex. In settings of disordered sinoatrial node impulse initiation or propagation that interfere with normal cardiac function, one has the choice of attempting to remake or repair the sinoatrial node—a difficult and as yet impossible task—or to reproduce aspects of its function. The overall goals are to restore the quality of life and to prolong life itself.

Pathologies Influencing Normal Pacemaker Function and a Brief History of Therapies

In 1827 and 1846, respectively, Adams¹ and Stokes² noted the clinical characteristics of an event that had likely afflicted mankind for centuries: They described patients who became pale, whose pulse became slow or was absent, and who then collapsed. We now recognize that the pathology was a high-degree heart block. In the 20th century, the availability of synthetic catecholamines and later of β-agonists afforded one form of treatment: sublingual isoproterenol every 2 hours.¹⁷ Although this therapy increased the rate of idioventricular pacemakers to a more physiological range in many patients, it also could and did cause ventricular tachycardia. Moreover, this stopgap therapy could not ensure any long-term survival.

The answer to this dilemma faced by patients and physicians was the electronic pacemaker. Discovery of this device is traced to the 1889 report of McWilliam, who described a means of delivering shocks at about 60 to 70 times per minute to a patient in whom the heartbeat had failed.¹⁸ In 1928, Lidwell and Booth used an electrical power source attached to a needle plunged into the heart to revive “a stillborn infant.”¹⁹ The term cardiac pacemaker is attributed to Hyman, who used a hand-cranked device to generate electrical shocks in 1932.²⁰ However, the modern era in pacing commenced in the early 1950s, when Hopps²¹ in Canada and then Zoll²² in the United States reported devices that delivered transcutaneous shocks. Pacing via shocks delivered through an intramyocardial needle in the setting of postsurgical complete heart block was reported by Weirich et al in 1957²³—the same year in which Bakken fabricated the first portable external pacemaker.²⁴

Initially, pacemaker implantation in the human heart required a thoracotomy for intramyocardial electrode placement.⁴ The first report of temporary transvenous approaches was published by Furman and Schwedel in 1959.³ The early 1960s then saw the rapid adoption of transvenous pacing with implanted units. Early pacemakers had no sensing function, paced at a fixed rate, and, although internally implanted, were cumbersome and required frequent power pack changes because of their use of mercury batteries. Moreover, their failure to sense often led to competition between implanted and idioventricular pacemakers and, in far too many instances, arrhythmogenesis. But the 1960s and the 1970s saw rapid improvement in batteries, electrodes, and software, such that pacemakers could be placed in a demand mode, sensing spontaneously occurring heartbeats and re-setting appropriately.^4,24,25

Advances continued through the 20th century and into the current decade, bringing AV sequential pacing that synchronized the sequence of atrial and ventricular activation in patients having normal sinus node function and AV block; exploration of the epicardial veins of the heart as sites for biventricular pacing to improve cardiac output; and the evolution of cardioverter-defibrillators that could sense, shock, and then pace if needed, for those patients who had potentially lethal tachyarrhythmias.

Innovation continues through the present time. For example, units are being tested that vary heart rate according to the body’s physiological needs,²⁶ and leadless electrodes are being developed that hold the promise of stimulating the heart without incorporating a catheter.²⁷ So the history of electronic pacing has been meteoric, and its future is bright.

Rationale for Developing Biological Pacemakers

Given the previous description of electronic pacemakers, why bother with biological pacing—or indeed with any other approach? Simply because the electronic pacemaker, although a wonderful treatment, is a panacea, not a cure. Table 26-1 summarizes the reasons why biological pacemakers are being investigated to normalize cardiac rhythm in a way that more nearly approximates the physiological state when compared with electronic pacemakers. If biological pacing succeeds, the two technologies would likely be used in tandem for some time to maximize patient safety. But the eventual goal would be to see the phasing out of electronic pacing. Therefore, biological pacing represents a disruptive technology.

Viral Vector-Based Delivery of Constructs

As can be seen in Figure 26-1, A, the initial ion channel-based approach reported transfection of guinea pig ventricles with an adenoviral vector incorporating a dominant negative Kir2.1 construct to reduce the repolarizing current, I_K1.³⁰ As is shown on electrocardiogram (ECG) in the intact heart and with microelectrode and ion current recordings, an ectopic ventricular rhythm was expressed that was associated with phase 4 depolarization and reduced I_K1 in myocytes. A shortcoming was that reducing repolarizing current increases action potential duration, and a subsequent publication showed that the repolarization characteristics of Anderson’s syndrome were created by this approach to biological pacing,³¹ such that its continued development was undesirable.

Our group has worked primarily with means to modify and magnify the pacemaker current, I_f, as originally reported by DiFrancesco.10,11 We first demonstrated in neonatal rat myocytes in cell culture that using an adenoviral vector to overexpress the HCN2 channel isoform resulted in a robust and rapid pacemaker that far exceeded in rate and stability the rhythm normally recorded in non-transfected cultures.^32,33 For this reason, we adopted the strategy of using HCN genes to build biological pacemakers and centered the approach on HCN2.³⁴ An additional attraction of working with HCN channels was that their inward current flows only during diastole; hence, although we were increasing an inward current, this would not prolong action potential duration, thereby avoiding the problem that had afflicted I_K1 downregulation.

Our initial in vivo study involved injecting adenoviral HCN2 and green fluorescent protein (GFP) constructs into canine left atria, and several days later using vagal stimulation to suppress sinus rhythm.³⁴ The result consisted of escape beats originating from the implant site. Cells disaggregated from this site showed I_f (actually I_HCN2) about 100 times greater in magnitude than that in native atrial myocytes. Additional vagal stimulation terminated firing of this pacemaker locus, consistent with parasympathetic control.

A follow-up study used a custom-designed electrode catheter to locate the left bundle branch system in dogs and to inject the same adenoviral-HCN2 complex into the bundle branch.³⁵ When the vagi were stimulated to terminate sinus rhythm and induce AV block, a rhythm emerged from the injection site. Microelectrode studies demonstrated spontaneous and rapid impulse initiation at these bundle branch sites. Subsequently, we found that in dogs with radiofrequency-induced complete heart block, a biological pacemaker inserted into the left bundle branch system drove the heart regularly and stably.³⁶ These rhythms were mapped to their bundle branch origins (Figure 26-2).

For the 2 weeks during which stable HCN2 expression occurred, about 70% of beats originated from the injection site. Basal rates averaged 50 to 60 beats/minute (bpm).³⁶ The remaining 30% of beats were provided by the demand electronic pacemaker, set at an escape rate of 45 bpm. A control group of saline-injected animals with electronic pacemakers, also set at 45 bpm, had 90% of beats electrically stimulated, with the remainder arising from various idioventricular sites. Finally, the biological pacemaker manifested an approximate 50% rate increase in response to catecholamine injection. Taking this together with our earlier report,³⁴ we concluded that both vagal and sympathetic (or at least β-adrenergic) modulation of biological pacing was possible.

Optimizing Pacemaker Gene Constructs

Attempts are ongoing to replicate or improve on the function of HCN2 using designer-based K channel genes,³⁷ mutant or chimeric HCN channels,^36,38,39 and combinations of two channels. We first studied a mutant HCN2 (E324A) that showed some enhanced catecholamine responsiveness, but only modest improvement over wild type HCN2.³⁶ We then designed a chimera, HCN212, whose transmembrane portion included the pore of HCN1 (which has more positive activation than HCN2) and the amino- and carboxy-termini (the latter incorporating the cAMP-binding site) of HCN2.³⁹