History The sinus node was discovered by Arthur Keith and his student Martin Flack in 1906, when they were working on a series of cardiac preparations from several species¹. In fact the sinus node was first observed in the mole’s heart and the famous Keith and Flack paper was published one year later in a journal that no longer exists². Because the sinus node is the ‘primum movens’, but also the ‘ultimum moriens’, its discovery did not definitely settle the debate on the question whether or not automaticity is an intrinsic feature of the heart, as appears from the famous review of Eyster and Meek, published as the first paper of the first issue of Physiological Reviews in 1921³. Automaticity yes, oscillation no Today we know that the sinus node displays automaticity, although none of its components has an oscillatory behaviour. In his monograph The Music of Life Denis Noble⁴ describes how difficult it was for him as a student to get access to what was a powerful computer at that time for modelling work on heart rhythm. “Mr. Noble, where is the oscillator in your equations. What is it that you expect to drive the rhythm?”⁴ was the question that remained unanswered. The answer is that the sinus node is not an oscillator. It is primarily the absence of the powerful inward rectifier current (I_K1) that allows the sinus node to be spontaneously active. In working cardiac muscle the resting membrane potential is almost identical to the equilibrium potential for K⁺, because the conductance of I_K1 is predominant during diastole. Automaticity and membrane currents The sinus node has no resting membrane potential. Still, full blockade of the L-type Ca²⁺ current (I_Ca-L) or the rapid delayed rectifier current (I_Kr) causes quiescence at about -30 mV at least in single sinus nodal myocytes^5,6. The large difference between the maximum diastolic potential (-65 mV) and this ‘resting potential’ creates a physiological condition in which membrane currents, which as stated above lack oscillatory behaviour themselves, are – and remain – out of phase and keep ‘each other going’. In the absence of I_Kr, the membrane cannot reach the negative potential required for activation of I_Ca-L and I_f (“the pacemaker current”). In the absence of I_Ca-L, there is no action potential and no activation of I_Kr. Ironically, absence of I_f does not abolish pacemaking, but the current is still important for tuning of heart rate by the autonomic nervous system. Reductionism: sinus node cells are no pars pro toto In an era of DNA-mania⁴ and preponderance of reductionism it is important to emphasize that isolated sinus node myocytes lack the feature of regularity⁷. An isolated sinus node or right atrium beats at almost perfectly constant cycle length. In situ, physiological influences create heart rate variability which is, however, not an intrinsic feature of the sinus node. The fact that an inhibitor of I_Kr will silence isolated sinus node cells and will also silence strips of sinus node tissue, if detached from the right atrium⁸, exemplifies that sinus node cells behave differently when isolated, when connected to each other (like in an isolated sinus node) or in the intact heart (when the sinus node is connected with the right atrium). This is not restricted to sinus node cells. It is caused by electrotonic interaction and it is observed in the ventricles as well, where action potentials are substantially shorter in intact tissue than in isolated myocytes⁹. It should be realized that computer models are based on data obtained in single cells and that the membrane currents involved are not necessarily identical in the intact organ. Functional inhomogeneity The sinus node displays “functional inhomogeneity”. Its basic elements have each a (different) intrinsic cycle length and a (different) responsiveness to ions, temperature, mechanoelectrical feedback, drugs and last but not least (neuro)hormones. Thus, changes in one of these parameters will induce pacemaker shifts within the nodal area¹⁰. Because the sinus node cells are connected with each other by connexins, there is one prevailing cycle length in the whole nodal area. Changes in cycle length, e.g. after an increase in norepinephrine, may result from a shortening in cycle length of a pacemaker area that was not determining the cycle length of the whole sinus node before the change in autonomic tone. The chronotropic response of the intact sinus node cannot be deduced from the chronotropic responses of its constituent cells. For changes directed at acceleration of heart rate, nodal areas with high responsiveness to the agent involved determine heart rate. For changes directed at deceleration of heart rate (as with vagal stimulation) areas with low responsiveness determine heart rate. Functional inhomogeneity is an important concept for the understanding of heart rate variability and “accentuated antagonism” of the autonomic nervous system.