Truly steady-state currents may flow through otherwise inactivating channels if membrane potential is maintained in a restricted range (“window”). These currents are commonly referred to as “window currents”. The window corresponds to the region of overlap between “activation” and “inactivation” curves, in which the probabilities of channels to be open and not inactivated are both greater than 0 at steady-state. A window is normally present in cardiac myocytes for I_CaL and, to a lesser extent for I_Na (Fig. 1), but it may be enhanced, or shifted to different potentials, by channel mutations (e.g. the Nav1.5 ΔKPQ mutation) [1].

When, during repolarization, membrane potential moves into the “window”, a fraction of channels may recover from inactivation and immediately reactivate. The resulting depolarization supports autoregenerative activation which, albeit concerning a small fraction of total channels, may substantially distort membrane potential course (early afterdepolarizations, EADs) and even trigger propagated activity [2, 3]. The I_Na window is narrow and more negative than the plateau phase in normal ventricular myocytes; thus, in the latter I_CaL window more likely accounts for EADs generation [4] (Fig. 1). I_Na window, somewhat more prominent in normal Purkinje myocytes [5, 6], may be widened in all cell types by Na⁺ channel mutations.

A further mechanism of I_Na enhancement during repolarization is provided by an increase in the rate of channel recovery from inactivation relative to deactivation one. This mechanism, observed in some NaV1.5 mutations, is referred to as “reactivation under non-equilibrium conditions” [7]. This is because the excess Na⁺ current can be observed only during dynamic changes of membrane potential (non-equilibrium). This is not a proper steady-state component and, notably, it can escape detection by the standard “voltage step” clamp protocols used to assess the functional phenotype of channel mutations.

Figure 2a shows that sustained I_Na, identified as current sensitive to blockade by tetrodotoxin (TTX), exists also at potentials positive to the I_Na window (−65 to −50 mV; shown in in Fig. 1a for the same cell type). This current is provided by a component named “late” Na⁺ current (I_NaL). I_NaL, much smaller than I_NaT in normal ventricular myocytes, may be enhanced by more than 3-fold under pathological conditions [8]. Single-channel analysis reveals that the channel gating underlying I_NaL is complex and, in addition to late “scattered” openings, it includes a “burst mode” not observed during the transient component of I_Na (I_NaT) [9, 10]. This led to hypothesize that I_NaL might be carried by channel variants other than the one prevailing in cardiac myocytes (NaV1.5), whose expression would be enhanced under pathological conditions (a contribution of NaV1.8 channels to I_NaL in mouse and rabbit cardiomyocytes has been indeed reported [11]). However, I_NaL with all its gating modes, can be detected by expression of Nav1.5 channels in cell lines (Fig. 2b), where no other Na⁺ channel isoforms are present [10, 12]. This proves that I_NaL can be carried by the same molecular entity accounting for I_NaT; according to this view, I_NaL enhancement reflects a gating abnormality. Observations crucial for the development of this interpretation came from single channel analysis of a mutation (ΔKPQ) associated with LQT3 syndrome [13]. The mutant channel had increased probability of burst openings during sustained depolarization, a behaviour interpreted as instability of the inactivated state [1, 13]. The interpretation of I_NaL enhancement as a gating abnormality of Nav1.5 channels may hold true also for acquired conditions, as proven in the case of heart failure [14‐16]. Nevertheless, we cannot rule out that changes in isoform expression may contribute in some of the diverse conditions in which I_NaL enhancement occurs.

Because of recovery from inactivation upon repolarization takes some time, I_NaT availability depends on the interval between excitations. This is true also for I_NaL, thus conferring to this current “reverse” rate-dependency [17], which is partial at physiological rates (approx 50 % reduction from quiescence to 120 b/min) [18].

Mutations prolonging action potential duration (APD) cause enhancement of plateau Na⁺ current through one or more of the mechanisms described above. For instance, all the above (widened voltage window, inactivation instability and accelerated recovery) may be operative in ΔKPQ mutants [1], but only non-equilibrium reactivation appears to account for the APD prolonging effect observed with the Il768V mutant [7].

I_NaL flows during repolarization and may directly affect its course. In normal canine ventricles I_NaL is differentially expressed across the wall (M-cells and Purkinje cells > subendocardial cells > subepicardial cells) [17, 32, 33]. Thus, it is conceivably a player in the physiological transmural repolarization gradient and in its rate-dependency in the dog (APD restitution) [34]. In normal ventricular myocytes, I_NaL inhibition by ranolazine (I_Na blocker with selectivity for I_NaL vs I_NaT) causes negligible APD changes. On the other hand, the remarkable effects of I_Kr blockade on APD, on its rate-depencency and, most importantly, on repolarization stability are all reversed by ranolazine [18, 35, 36]. These apparently contrasting findings may be reconciled by considering that ranolazine also partially blocks I_Kr [37]. Under basal conditions, this may offset the effect of I_NaL inhibition on APD; in turn, concomitant I_NaL inhibition limits the effect of I_Kr inhibition, thus preventing repolarization instability. If this interpretation is correct, we can conclude that I_NaL and I_Kr are physiologically in balance during normal repolarization; whenever this balance is altered, by either I_NaL enhancement or I_Kr blockade, repolarization stability is compromised. An extreme example of this condition is advanced heart failure, in which I_NaL enhancement and I_Kr downregulation coexist and are associated with dramatic repolarization instability [15]. Although the concept of I_Kr - I_NaL balance is valid in a broad sense, the effects of I_Kr blockade and I_NaL enhancement on action potential contour are not identical, an observation which may have its counterpart in the differences of clinical presentation between the LQT2 (I_Kr deficiency) and LQT3 (I_NaL enhancement) syndromes [38].

The direct contribution of I_NaL to repolarization course provides a first powerful mechanism linking arrhythmogenesis to I_NaL enhancement. Nevertheless, the latter may also facilitate arrhythmias through Ca²⁺ handling abnormalities (see below) and it is difficult to establish which mechanism prevails in a specific condition. The mutual interplay between Ca²⁺ handling and repolarization course [39] may actually make this distinction pointless.

A role of I_NaL in arrhythmogenesis is often inferred from the antiarrhythmic effect of its selective blocker ranolazine. While this may be considered legitimate in many cases, there are exceptions due to specificities of drug action. The best example is ranolazine efficacy on atrial arrhythmias, to which mechanisms other than I_NaL inhibition may also contribute [40].

I_Na represents the main source of Na⁺ entry during the cardiac cycle. Abeit I_NaL amplitude is normally 1/1000 of that of I_NaT [10], I_NaL persists throughout repolarization; as a result, I_NaT and I_NaL may similarly contribute to Na⁺ influx during a cardiac cycle [41]. Nevertheless, I_NaL inhibition (by TTX or ranolazine) marginally affects Ca²⁺ cycling [42] and contractility [43] in normal hearts, thus suggesting that the attending changes in Na⁺ influx are effectively buffered by matching changes in Na⁺ extrusion. However, cellular homeostasis can be compromised by the marked increase in Na⁺ influx resulting from pathological I_NaL enhancement.

Na⁺ is normally extruded from the cell by the ATP powered Na⁺/K⁺ pump. Therefore excess Na⁺ influx, even when successfully buffered, may increase ATP consumption. Furthermore, if influx exceeds the maximal extrusion rate, Na⁺ accumulates in the cytosol, thus partially dissipating its transmembrane gradient. Because the latter energizes many secondary membrane transport mechanisms, most importantly the Na⁺/Ca²⁺ exchanger (NCX) and the Na⁺/H⁺ exchanger (NHE), a pivotal consequence of I_NaL enhancement is perturbed homeostasis of intracellular Ca²⁺ and H⁺ (Fig. 3).

NCX is the main mechanism of Ca²⁺ extrusion from the cell. Increased cytosolic Na⁺ moves its electrochemical equilibrium potential in the negative direction, thus reducing the driving force for its forward operation during diastole and possibly reversing the direction of transport during systole (i.e. Ca²⁺ entry through NCX). The resulting increase in intracellular Ca²⁺ may re-establish NCX driving force, but the system equilibrium is now moved to higher cytosolic Ca²⁺ levels. Under conditions of I_NaL enhancement (e.g heart failure) NCX expression may be upregulated [44] and, provided that a driving force still exists, this may increase Ca²⁺ transport rate. However, the effect of this change in sustaining Ca²⁺ extrusion can only be partial, because it vanishes as NCX electrochemical equilibrium is approached. Accordingly, unless maximal Na⁺/K⁺ pump transport rate is also incremented, an increase in total cellular Ca²⁺ content is a necessary consequence of I_NaL enhancement. A further aspect of interest is the distribution of such an increment between subcellular compartments. Stimulation of Ca²⁺ uptake by the sarcoplasmic reticulum (SR) (by SERCA2) is a central component of physiological stimuli meant to increase cell Ca²⁺ content (e.g. by β-adrenergic activation). This ensures that most of the gain concerns SR luminal Ca²⁺, which results in larger amplitude of Ca²⁺ transient (larger developed force) and lower diastolic Ca²⁺ (accelerated relaxation). This is not the case for I_NaL enhancement, in which the Ca²⁺ increment concerns primarily the cytosol, which may only secondarily increase Ca²⁺ in the SR. Although the latter may increase maximal developed force, this is expectedly associated with higher diastolic Ca²⁺. The consequences of cytosolic Ca²⁺ being persistently elevated are multiple and of pathophysiological relevance.

Opening of SR Ca²⁺ release channels (RyRs) is directly triggered by cytosolic Ca²⁺, a process facilitated by high Ca²⁺ in the SR lumen [45]. Increased cytosolic Ca²⁺ also activates CaMKII-dependent RyRs phosphorylation, a further mechanism of RyRs facilitation (see below, Fig. 3). It is therefore unsurprising that I_NaL enhancement may lead to facilitation of diastolic “Ca²⁺ waves” and the resulting electrical disturbances (delayed afterdepolarizations, DADs) [46]. This represents an important arrhythmogenic mechanism, probably the one prevailing under conditions of altered Ca²⁺ handling (e.g. heart failure).

With persistently elevated cytosolic Ca²⁺, diastolic function may be hampered by delayed and incomplete sarcomere relaxation. A contribution of I_NaL to diastolic dysfunction has been demonstrated in various experimental models of heart failure [25, 42] and ranolazine improved diastolic relaxation in human ischemic heart disease [47]. In addition to its direct hemodynamic impact, diastolic dysfunction may limit coronary flow by extrinsic compression of intramural vessels. Significance of this effect is indirectly demonstrated by the ability of ranolazine to improve myocardial perfusion in the setting of coronary artery disease (likely I_NaL enhancement) [48]. Indeed, being ranolazine devoid of significant direct vasodilator effects, the perfusion improvement is likely to result from accelerated diastolic relaxation, a well known factor in modulation of coronary flow [49].

Abnormally elevated cytosolic Ca²⁺ may also activate a number of signalling pathways involved in modulation of cell function and, in the long run, structure and fate. CaMKIIδ activation, is at the same time, a cause (see above) and a consequence [50] of I_NaL enhancement. Phosphorylation of SR Ca²⁺ release channels (RyRs) by this kinase increases their open probability [51], ultimately leading to destabilization of the Ca²⁺ store. This accounts for the facilitation of spontaneous Ca²⁺ release events and delayed afterpotentials (DADs, their arrhythmogenic electrical consequence) induced by either I_NaL enhancement or CaMKIIδ overexpression [52] and suppressed in both cases by I_NaL blockade [25]. Calcineurin, a cytosolic Ca²⁺-activated phosphatase, regulates translocation to the nucleus of NFAT, a transcriptional regulator centrally involved in hypertrophic remodelling [53]. While there is still no direct evidence for the involvement of this specific pathway in I_NaL-induced damage, ranolazine has been shown to affect downstream gene transcription and histomorphometric parameters in heart failure induced by coronary microembolization [54].

A further consequence of pathological I_NaL enhancement is a derangement in cell energy balance. The main ATP-consuming mechanism involved in the control of cellular homeostasis (the Na⁺/K⁺ and SERCA pumps) are conceivably short-circuited by enhanced sarcolemmal Na⁺ influx and increased Ca²⁺ leak from the SR (by RyRs facilitation) (Fig. 3). Therefore, I_NaL enhancement is expected to increase ATP consumption by mechanisms not directly involved in force generation, which would translate into decreased mechanical efficiency. This view is supported by the observation that I_NaL blockade prevents the drop in myocardial ATP, but not the positive inotropic effect, of ouabain [55]. In addition to improved coronary perfusion, increased myocardial efficiency might partly account for the ability of I_NaL blockade to improve exercise capacity of ischemic patients without altering systolic mechanical work [56].

It has been reported that increased cytosolic Na⁺ may also jeopardize mitochondrial function (Fig. 4). Concomitant cytosolic Na⁺ loading (to 15 mM) impaired Ca²⁺-triggered mitochondrial energetic adaptation in patch-clamped myocytes, thereby causing an abnormal drop in NADH (i.e metabolic energy) upon catecholamine challenge [57]. On the other hand, ranolazine pre-treatment failed to modify NADH course during global ischemia/reperfusion in isolated hearts [43]. Notably, in the same preparation, ranolazine opposed ischemia-induced rise in cytosolic and mitochondrial Ca²⁺; this was associated, as expected, with diminished ROS generation and delayed opening of mitochondrial permeability transition pore (mPTP) [43]. A plausible network of mechanisms in accord with these observations is illustrated in Fig. 4. Altogether, this preliminary evidence suggests that, during ischemia/reperfusion, I_NaL inhibition may be more effective in preserving integrity of mitochondria than in improving their function. Although ranolazine effects on energy metabolism can be accounted for by I_NaL inhibition, the drug has been also reported to inhibit fatty acid oxidation [58, 59], a further action potentially protecting mitochondria during metabolic stress. However, this action was observed with drug dosage (60–200 mg/Kg/day) far from that required for the anti-ischemic effect in humans (10–15 mg/Kg/day).

Under normal conditions, cellular H⁺ homeostasis may be less sensitive than Ca²⁺ homeostasis to I_NaL enhancement, because mechanisms other than NHE contribute to H⁺ extrusion. During abrupt reperfusion following acute coronary occlusion, massive H⁺ extrusion through NHE contributes to intracellular Na⁺ loading, which explains why direct NHE inhibition paradoxically improves Ca²⁺ handling and limits myocardial injury [60]. Nevertheless, the situation might be different in the presence of chronic partial ischemia, when reduced H⁺ export by NHE might lead to persistent intracellular acidosis, potentially contributing to impair force development and relaxation. Although awaiting experimental confirmation, this hypothesis is supported by the observation that the increment in intracellular Na⁺ caused by Na⁺/K⁺ pump blockade does produce intracellular acidosis, which can be prevented by concomitant I_NaL blockade [55].