Pharmacologic Approach to Sinoatrial Node Dysfunction

Abstract

The spontaneous activity of the sinoatrial node initiates the heartbeat. Sinoatrial node dysfunction (SND) and sick sinoatrial syndrome are caused by the heart’s inability to generate a normal sinoatrial node action potential. In clinical practice, SND is generally considered an age-related pathology, secondary to degenerative fibrosis of the heart pacemaker tissue. However, other forms of SND exist, including idiopathic primary SND, which is genetic, and forms that are secondary to cardiovascular or systemic disease. The incidence of SND in the general population is expected to increase over the next half century, boosting the need to implant electronic pacemakers. During the last two decades, our knowledge of sinoatrial node physiology and of the pathophysiological mechanisms underlying SND has advanced considerably. This review summarizes the current knowledge about SND mechanisms and discusses the possibility of introducing new pharmacologic therapies for treating SND.

INTRODUCTION

The cardiac impulse is generated in the sinoatrial node (SAN) by a highly integrated mechanism involving ion channels, intracellular Ca²⁺ dynamics, membrane receptors, and connexins (1, 2). Despite the intrinsic robustness of the pacemaker mechanism, SAN dysfunction (SND) constitutes a relatively common clinical condition, especially among people over age 65 (1/600) (3). SND generally manifests as an insufficiency of the heart rate to meet the needs of the organism. SND may be differentiated into reversible (acute) or chronic symptomatic forms (4).

Acute SND is generally treated pharmacologically or by temporary transvenous or transesophageal pacing (4, 5). In contrast, only a few options are currently available to treat chronic symptomatic SND, and to date implantation of an electronic permanent pacemaker (PPM) remains the primary and definitive therapy (4, 5). In this regard, symptomatic SND and atrioventricular block account for approximately half of the total pacemaker implantations in the United States (6), and the number of pacemaker implantations is predicted to double over the next fifty years (7). In addition, clinical studies indicate an increasing necessity for implantation of complex electronic pacemakers (8).

In this review, we discuss some of the most documented forms of primary and secondary SND, especially in relation to current pharmacologic management. Research over the last 20 years has considerably advanced our knowledge of SND etiology. Several genes coding for ion channels (9–21), scaffolding proteins (22), and cytoskeleton proteins (23), as well as connexins and proteins involved in cardiac development have been linked to previously unexplained forms of primary familial SND (24). Furthermore, development of animal models of primary and secondary SND, as well as ex vivo exploration of SANs from human hearts with a history of SND, have shed new light on the mechanisms of secondary forms of SND. These approaches have identified several novel targets for managing SND, including cardiac G protein–activated I_KACh/I_KAdo (GIRK1/GIRK4) (25–29) and small-conductance Ca²⁺-activated K⁺ channels (30–32). The identification of GIRK channels as potential molecular targets for SND may extend the indication of pharmacologic approaches to chronic forms of SND. More generally, it is possible that innovative molecules targeting specific SND mechanisms will help manage chronic SND, which is now treated only with PPMs.

CLINICAL DEFINITION OF SINOATRIAL NODE DYSFUNCTION

The diagnosis of SND is based on the correlation between the patient’s symptoms and electrocardiogram (ECG) hallmarks (see also Supplemental Appendix 1), which provide important criteria for PPM implantation (4, 5, 33). Historically, patients with SND have been identified as having one or more of the following findings from ECGs (34): (a) persistent, unexpected sinoatrial bradycardia (Figure 1); (b) short periods of sinoatrial arrest during which atrial or junctional rhythms replace normal sinoatrial rhythm; (c) long periods of sinoatrial arrest in the absence of junctional rhythms, resulting in cardiac standstill; and (d) episodes of conduction block between the sinoatrial node and the atria (sinoatrial exit block) not related to drug therapy (34, 35). This early definition of SND remains in use in current clinical guidelines (4, 5, 36). Sinoatrial bradycardia is generally defined as a heart rate below 50 beats per minute (bpm) (4). Sinoatrial pauses or sinoatrial arrest is included in the current definition of SND, particularly when manifest as tachycardia-bradycardia syndromes, in which sinoatrial bradycardia, pauses, or arrest follows periods of abnormal atrial tachycardia, atrial fibrillation, or flutter (37). In tachycardia-bradycardia syndromes, SND can manifest as poor or sluggish return of sinoatrial rhythm following cardioversion (38) (Supplemental Figure 1). Another hallmark of SND is chronotropic incompetence, defined as the inability of the heart rate to attain 80% of the expected heart rate during exercise (39). SND can carry severe symptoms, seriously compromising the patient’s quality of life (36). One of the most common symptoms of SND is syncope, which is present in approximately half of patients with SND (5, 40). Whereas asymptomatic bradycardia is not associated with adverse outcomes, patients with untreated symptomatic SND have a high risk of deterioration to cardiovascular events including atrial fibrillation (41), heart failure (42), and systemic thromboembolism (5, 40). Age-dependent SND and chronotropic incompetence are associated with an increased risk of cardiovascular death and overall mortality (4).

Figure 1.

A 12-lead ECG showing severe sinoatrial bradycardia and a rhythm strip showing sinoatrial pauses as manifestations of SND. The baseline heart rate was 45 beats per minute. Figure adapted with permission from Dr. Francis Marchlinski, University of Pennsylvania. Abbreviations: ECG, electrocardiogram; SND, sinoatrial node dysfunction.

SINOATRIAL NODE PACEMAKING: GENERAL OVERVIEW

The SAN is a highly complex, heterogeneous tissue (1). Surprisingly, pacemaker cells do not constitute the predominant cell type in the SAN. Early studies of SAN tissue indicated that atrial myocytes (43) and fibroblasts (44) are important constituents of SAN structure and integrative properties. In 2020, a single-cell RNA sequencing–based study indicated that the SAN is composed of atrial myocytes, adipocytes, epithelial cells, fibroblasts, vascular endothelial cells, macrophages, and neurons (45). In addition to heterogeneous cellular composition, pacemaker cells within the SAN are poorly coupled electrically (46). This low intercellular conductance is due to high expression of connexin45 (Cx45) and low or absent expression of Cx43 in pacemaker cells (46–48). Finally, the SAN region includes specific nonconductive structures and redundant impulse propagation pathways (29, 49) that help ensure proper intrasinoatrial conduction and sinoatrial-to-atria conduction (50, 51).

Normal pacemaking depends on a unique action potential (AP) profile of SAN pacemaker cells. Importantly, the SAN AP undergoes a spontaneous diastolic depolarization phase, driving the membrane voltage from the end of the repolarization to the threshold of the following AP (Figure 2). Catecholamines positively regulate the slope of the diastolic depolarization via activation of β-adrenergic receptors (βARs). The adrenergic activation stimulates the synthesis of cyclic adenosine monophosphate (cAMP), which positively regulates the activity of several ion channels of the plasma membrane, and the intracellular ryanodine receptors (RyRs)/Ca²⁺-release channels embedded in the sarcoplasmic reticulum (SR) (2, 52) (Figure 2). In antagonism with βARs, muscarinic type 2 receptors and adenosine type 1 receptors decrease pacemaking by promoting downregulation of intracellular cAMP and by inducing the opening of G protein–activated K⁺ channels (GIRK1/GIRK4) underlying the I_KACh/I_KAdo current (53). The current understanding of how the heartbeat is generated is best summarized by the coupled-clock model of pacemaking (52). The coupled-clock model states that diastolic depolarization is generated by the functional interplay between activity of ion channels at the plasma membrane and local diastolic type 2 ryanodine receptor (RyR2)-dependent Ca²⁺ release, which is coupled to diastolic depolarization via activation of the cardiac Na⁺/Ca²⁺ exchanger (NCX1) (2, 52, 54). Different ion channels of the plasma membrane contribute to the generation and the regulation of diastolic depolarization and are discussed here briefly.

Figure 2.

(a) Schematic of a SAN pacemaker cell showing the main membrane-bound receptors and ion channels, together with RyR2, SERCA, and NCX1. (b) Simulated transmembrane potential (V_m) during a spontaneous AP generated by a computational model of the mouse SAN myocyte (177), with colors corresponding to the major AP phases. Colors also indicate corresponding underlying currents in panel a: diastolic depolarization (red), AP upstroke (green), and repolarization (blue). See also the extended version of this figure (Supplemental Figure 2). Abbreviations: βAR, β adrenergic receptor; A1, type 1 adenosine receptor; AC, adenylate cyclase; AP, action potential; Ca_v1.2, L-type Ca_v1.2 Ca²⁺ channel; Ca_v1.3, L-type Ca_v1.3 Ca²⁺ channel; Ca_v3.1, T-type Ca_v3.1 Ca²⁺ channel; ERG1, type 1 cardiac ether-a-gogo related delayed rectifier K⁺ channel; Gα_i, G protein α_i subunit; Gβ_γ, G protein β_γ subunits; GIRK, G protein–activated K⁺ channel; HCN, hyperpolarization-activated cyclic nucleotide-gated channel; I_st, sustained Na⁺ current; M2, type 2 muscarinic receptor; Na_v1.1, voltage-dependent TTX-sensitive Na⁺ channel; NCX1, Na⁺/Ca²⁺ exchanger 1; RyR2, type 2 ryanodine receptor; SAN, sinoatrial node; SERCA, sarco-/endoplasmic reticulum Ca²⁺ ATPase; SR, sarcoplasmic reticulum.

The hyperpolarization-activated funny current (I_f) is activated at the end of the repolarization phase of the AP and supplies inward current throughout the range of diastolic depolarization (55). Catecholamines shift the I_f activation curve to more positive voltages, whereas acetylcholine induces a negative shift. These opposing effects are explained by the direct sensitivity of f-channels to cAMP (56), which increases the probability of channel opening at a given voltage (57). The hyperpolarization-activated cyclic nucleotide-gated channel (HCN) family, which comprises four distinct isoforms HCN1-HCN4, encodes the f-channels. However, in the SAN the predominant isoform is HCN4, which accounts for 80% of the total HCN messenger RNAs (mRNAs) (58). Moreover, the HCN1 protein is almost exclusively expressed in the human SAN rather than in the atrial myocardium (59). In the SAN, voltage-gated Ca²⁺ channels of the Ca_v gene families contribute both to diastolic depolarization and to AP upstroke. SAN pacemaker cells express T- and L-type Ca_v channels, which underlie voltage-dependent I_CaT and I_CaL Ca²⁺ currents, respectively. Both Ca_v3.1 and Ca_v3.2 mRNAs are expressed in the SAN; however, the predominant functional T-type isoform in the adult SAN is Ca_v3.1 (60). In spite of this low availability, Ca_v3.1 knockout (Ca_v3.1^−/−) mice present a moderate reduction in SAN rate (−10%) and prolonged atrioventricular conduction interval (PR interval) (60). SAN pacemaker cells concomitantly express two distinct L-type Ca²⁺ channel isoforms, Ca_v1.3 and Ca_v1.2 (61). Ca_v1.3-mediated I_CaL is characterized by a more negative threshold for activation than Ca_v1.2-mediated I_CaL (−45 mV for Ca_v1.3-mediated I_CaL versus −25 mV for Ca_v1.2-mediated I_CaL) (61). Under basal conditions, pacemaker cells from Ca_v1.3^−/− mice show erratic generation of automaticity, a lack of linear phase of the diastolic depolarization (62), and a strong reduction of the total inward diastolic current compared with pacemaker cells from control littermates (26, 62).

The SAN expresses voltage-gated fast tetrodotoxin (TTX)-sensitive (I_NaTTX) and TTX-insensitive Na⁺ (I_Na) currents. I_NaTTX is encoded by neuronal (n)Na_v Na⁺ channel isoforms and contributes to pacemaking and intranodal conduction in the mouse, rabbit, and human SAN (63–66). The nNa_v1.1 and nNa_v1.3 isoforms have been proposed to contribute to sinoatrial I_NaTTX in mouse and rabbit (63), whereas nNa_v1.6 appears to play a dominant role in human intranodal conduction and SND (66). The cardiac (c)Na_v1.5 isoform underlies the I_Na in the mouse SAN. cNa_v1.5 channels contribute to intranodal or nodal-atrial impulse conduction (65–67).

Ion channels of the transient receptor potential (TRP) channel family contribute to SAN activity (68–70). TRPC channels contribute to store-operated Ca²⁺ entry in pacemaker cells (70). TRPM7 (TRP cation channel subfamily M member 7) channels contribute to pacemaking by regulating expression of HCN4 channels in the sinoatrial and atrioventricular nodes (71). TRPM4 channels contribute to the basal beating rate of SAN pacemaker cells (68).

In the last decade, members of the Ca²⁺-activated K⁺ channel family (K_Ca) underlying I_KCa have been linked to pacemaking in mouse and rabbit SAN. K_Ca channels have been subdivided into big- (72), intermediate- (31), and small-conductance (32, 73) K⁺ channels. These channels have differential sensitivity to Ca²⁺. Big-conductance K_Ca (BK) channels are primarily voltage dependent with positive regulation by Ca²⁺ (74). In contrast, intermediate- and small-conductance K_Ca (IK and SK, respectively) channels are voltage independent and highly sensitive to Ca²⁺ (31). Such sensitivity to Ca²⁺ is mediated by the association with calmodulin (CaM) (74) and, at least for SK channels, coupling with L-type Ca²⁺ channels has been demonstrated (75, 76). SK isoforms have different sensitivities to the specific inhibitor apamin (SK1, SK2, and SK3) (74). All three SK isoforms are expressed in the SAN (32).

FAMILIAL PRIMARY SINOATRIAL NODE DISEASE AND ASSOCIATED CONDUCTION DEFECTS

Although SAN disease and SND are often associated with aging or with different cardiovascular pathologies such as heart failure (77, 78), work over the past two decades has identified familial forms of SND (24, 79). In fact, congenital forms of SAN disease have been instrumental in providing molecular insights into the critical and nonredundant molecular pathways that underlie SAN automaticity and signaling. Multiple ion channels and ion channel regulatory subunits, as well as structural proteins, are now associated with SAN disease and dysfunction in impulse conduction and are discussed here (Supplemental Table 1).

HCN4 variants are linked to human asymptomatic bradycardia or SAN disease (9–12). Mechanistically, variants have been associated with altered channel membrane targeting, aberrant ion channel activation (10), or conductance (12), as well as dysfunction in channel regulation by cyclic nucleotides (9, 13). Of note, human HCN4 variants are linked to other forms of cardiovascular disease beyond SND, demonstrating the key role of this ion channel (11). Human variants in SCN5A, which encodes cNa_v1.5 channels, are linked to progressive disorders of cardiac conduction such as Lev-Lenègre syndrome (80). More generally, loss of function of SCN5A accounts for 5% of the total incidence of primary conduction system dysfunction in humans (21, 81). SCN5A variants are linked to familial bradycardia, conduction disorders, and atrioventricular block (24, 79). In addition, some SCN5A mutations are also linked to SND (20, 82; see 83 for review). This finding is consistent with evidence showing that high concentrations of TTX (1–3 μM) inhibiting cNa_v1.5 channels decrease the intranodal conduction time and induce exit block in intact human SAN (66). However, similar to variants in other key cardiac ion channels, SCN5A variants are linked to other forms of atrial and ventricular arrhythmia (84).

Ca_v1.3 channels are encoded by CACNA1D. Although not expressed in the working myocardium as robustly as Ca_v1.2 channels, Ca_v1.3 channels are highly expressed in the SAN (61, 85–87) (see also the section titled Sinoatrial Node Pacemaking: General Overview). Though rare, human CACNA1D variants that alter channel activity are linked to human SND (14, 15). Ca_v1.3 channels have also been linked to neonatal complete atrioventricular block related to lupus (88). Mechanistically, atrioventricular block has been explained by the presence of maternal autoantibodies against Ca_v1.3 channels (89, 90). In addition to Ca_v1.3 channels, T-type Ca_v3.1 channels (encoded by CACNA1G) have also been linked to human bradycardia and congenital complete heart block related to maternal lupus (88, 91).

TRPM4 is a nonselective Ca²⁺-regulated channel expressed in the SAN and atria. TRPM4 variants have been widely linked to numerous human cardiac phenotypes, including sinoatrial bradycardia, likely through modulation of the SAN cell membrane potential (92, 93).

Alterations in several SAN and atrial accessory and calcium regulatory proteins have been associated with human SND. Ankyrin-B, encoded by ANK2, is a cytoskeletal adaptor protein that associates with cardiac ion channels, transporters, signaling molecules, and structural proteins. Consistent with animal models lacking ankyrin-B, humans harboring specific ANK2 variants may display bradycardia and conduction defects (22). Mechanistically, ankyrin-B dysfunction alters multiple critical SAN proteins, including Ca_v1.3 and NCX1, resulting in aberrant diastolic depolarization. The influence of specific genetic and environmental factors on SND penetrance and severity in the human population is based on the degree of ankyrin-B loss of function and potential secondary variants or environmental factors. Similar to ankyrin-B, cardiac caveolin-3 (CAV3) variants, although linked to ventricular arrhythmia, are also associated with bradycardia (94). Mechanistically, CAV3 variants may affect multiple ionic currents to influence SAN automaticity and particularly cNa_v1.5 and nNa_v1.x channels.

During the last five years, an exciting set of papers have linked human variants in β-subunits of heterotrimeric G proteins to SND. To date, both GNB5 and GNB2 variants are linked to sinoatrial bradycardia and SND (17, 18). Individuals harboring GNB variants may display additional noncardiac phenotypes, including cognitive disorders. Mechanistically, these variants may affect the activity of I_KACh, inducing current gain of function and consequent sinoatrial bradycardia and atrioventricular block. Consistent with this hypothesis, mutations in KCNJ3 and KCNJ5 inducing gain of function of GIRK1 and GIRK4 were linked to familial SAN disease (KCNJ3 encodes GIRK1; KCNJ5 encodes GIRK4) (16, 95).

Dysfunction in cardiac calcium regulatory proteins are linked to familial SND. Calsequestrin 2 (CASQ2) is a calcium-binding protein that also serves a critical role in regulating local calcium release and automaticity. The cardiac gene RYR2 is a central regulatory node responsible for excitation–contraction coupling. Variants in either RYR2 or CASQ2 are associated with bradycardia and potentially fatal arrhythmias in response to catecholamines (96–98). In particular, catecholaminergic polymorphic ventricular tachycardia (CPVT) is a life-threatening familial ventricular arrhythmia associated with mutations in CASQ2 or RYR2 (99). RYR2-associated CPVT is characterized by increased Ca²⁺ release from the SR at rest and under adrenergic activation (99). However, some patients with CPVT who carry mutations in RYR2 also present moderate sinoatrial bradycardia (100, 101). Bradycardia has been explained by tonic Ca²⁺-dependent inactivation of I_CaL and reduction of basal SR Ca²⁺ load in SAN pacemaker cells (100, 101).

Additional forms of inherited SND involve cardiac proteins with a range of functions in the myocyte. For example, human variants in MYH6, which encodes cardiac muscle myosin, are linked to SND (as well as other non-SAN electrical and structural phenotypes) (102). Like myosin, lamins A and C have multiple roles for cardiac function. Encoded by LMNA, lamin A variants are associated with sinoatrial bradycardia and conduction system disorders (103). Finally, variants in numerous other genes, including the transcription factor–encoding SHOX2 (104), involved in differentiating the SAN, have been linked to cardiac SAN and atrioventricular block phenotypes.

Like other forms of familial human disease, disease penetrance and disease severity depend on secondary genetic, environmental, and social factors. Thus, caution should always be utilized when interpreting genetic information from individual variant carriers.

GENETIC MOUSE MODELS OF PRIMARY SINOATRIAL NODE DYSFUNCTION

A wide variety of genetically modified mouse models have been generated to study SAN disease. Among the first mouse models of SND were those involving knockout of ion channels important for SAN membrane excitability. Early studies found that Ca_v1.3^−/− mice showed congenital deafness, bradycardia, and irregular heart rate (105). Further work has shown that Ca_v1.3^−/− mice constitute a model of the SAN dysfunction and deafness (SANDD) syndrome, which is characterized by sinoatrial bradycardia and atrioventricular conduction dysfunction (14). Indeed, Ca_v1.3^−/− mice have pronounced sinoatrial bradycardia, associated with sinoatrial pauses, atrial fibrillation, and flutter, and second-degree atrioventricular block (26, 61, 106). Global knockout of Hcn4 prevents proper development of the SAN and conduction system, resulting in embryonic lethality (107). An inducible Hcn4 knockout mouse model has also been generated and has either mild SND with sinoatrial pauses (108) or severe bradycardia and conduction system defects incompatible with life (109). Knockout of Hcn1 induces bradycardia and SND in the mouse (110). Mice carrying genetically silenced I_f conductance show sinoatrial bradycardia and SND associated with second-degree atrioventricular block and ventricular arrhythmia (25). By contrast, mice lacking cAMP-dependent regulation of Hcn4 show moderate sinoatrial bradycardia, consistent with that observed in humans carrying cAMP-regulation-defective HCN4 (9, 13). Similar to clinical findings in studies of humans, mice haplo-insufficient for cNa_v1.5 (cNa_v1.5^+/−) display lack of P waves, consistent with sinoatrial exit block, reduced excitability of pacemaker cells, and intraventricular conduction defects (67).

Researchers use genetic mouse models to explore proteins important for intracellular ion homeostasis. Knockout of atrial Ncx1 (sarcolipin-Cre crossbred with Ncx1 floxed mouse) led to sinoatrial arrest, burst pacemaking, and junctional escape rhythm (111). In a similar vein, knockout of the SR Ca²⁺ buffer Casq2 disrupted SAN Ca²⁺ handling, leading to irregular pacemaking and SND (112), consistent with observations in patients (98).

There is a growing number of mouse models with SND and defects in unconventional ion channels and accessory proteins not typically invoked in the context of SAN pacemaking. For example, the background K⁺ channel TREK-1 emerged in 2016 as a novel determinant of SAN excitability and pacemaking. Specifically, mice with cardiac specific ablation of TREK-1 show bradycardia with frequent sinoatrial pause (113). At the cell level, TREK-1-deficient SAN pacemaker cells show a depolarized maximum diastolic potential and altered firing rate. Similarly, several TREK-1-interacting partners have been linked to SND in mice, including the cytoskeletal protein βIV-spectrin (114) and members of the Popeye-domain-containing (POPDC) gene family (115). Mouse models of TRPM4-related SND have been generated. Although the TRPM4 knockout (Trpm4^−/−) mouse shows no difference in heart rhythm compared with wild-type counterparts, pharmacologic inhibition of TRPM4 slows spontaneous beating in wild-type but not Trpm4^−/− atria (116). Furthermore, TRPM4 mice show more frequent episodes of sinoatrial pauses and prominent conduction block compared with wild-type counterparts (117). TRPM7 is a divalent channel kinase abundantly expressed in mouse and human heart. Global as well as sinoatrial/atrioventricular node–restricted knockout of TRPM7 slows diastolic depolarization in pacemaker cells and induces bouts of sinoatrial pause and conduction block (71). As discussed above, ankyrin-B is an adapter protein important for proper membrane localization of multiple ion channels and transporters for normal SAN excitability and pacemaking. Human variants in ANK2 have been linked to a complex arrhythmia syndrome that includes severe SND and requires pacemaker implantation (22). Consistent with the human phenotype, mice heterozygous for ankyrin-B display severe bradycardia and enhanced resting heart rate variability compared with wild-type littermates (22).

Mouse models have been instrumental in studying the link between dysregulated gene expression and SND. Notably, mice deficient for the transcription factor Tbx3 show defects in development of the SAN and of the conduction system. They also show SND characterized by bradycardia and sinoatrial pause presenting in the embryonic stage (118). The cardiac homeobox transcription factor Nkx2–5 has been studied with an atria-specific knockout model, which shows hyperplasia of the working myocardium and conduction system, resulting in a broad spectrum of arrhythmias including bradycardia, conduction block, perinatal lethality (119). Mechanistically, Nkx2–5 deficiency activates Notch signaling and enhances myocyte proliferation early in development. Notch signaling is also activated by injury and a genetic mouse model has been used to study the effects of transient Notch activation on the atria (120). Notch activation reduced the expression of cNa_v1.5 and induced structural remodeling of the SAN, resulting in sinoatrial bradycardia and sinoatrial pause (120). Genetic mouse models have also elucidated a role for signaling through the multifunctional Ca²⁺/CaM-dependent kinase II (CaMKII) in SND in the setting of neurohumoral dysregulation. Specifically, transgenic mice overexpressing a CaMKII inhibitory peptide (AC3-I) were resistant to development of fibrosis and sinoatrial pause following chronic angiotensin II infusion (121). As new experimental models have been developed and applied to study SAN function and disease, so too have numerous computational models of pacemaker activity (see Supplemental Appendix 2). These models will prove important for understanding SND as the discovery of new mechanisms progresses.

SECONDARY FORMS OF SINOATRIAL NODE DYSFUNCTION IN HUMANS AND ANIMAL MODELS

Many forms of SND are associated with a variety of systemic diseases. When SND presents in association with a systemic disease, or can be a consequence of such a disease, the term secondary or acquired SND is used. In fact, secondary forms of SND account for most patients. The incidence of SND correlates with age and comorbidities such as hypertension, diabetes, presence of cardiovascular disease, and elevated plasma concentrations of cystatin-c and natriuretic peptide (7). However, current guidelines generally make a distinction between secondary forms of SND associated with systemic disease, cardiovascular disease, and drug intoxication (4). In relation to systemic conditions, SND can be a secondary manifestation of endocrine disease such as hypothyroidism and diabetes, of inflammatory or rheumatologic disorders, of plasma ionic imbalance, or of infectious disorders (4, 77).

Abnormal input by the autonomic nervous system is a potential cause underlying SND. In some patients, intrinsic SND can be worsened by autonomic nervous system imbalance (122). In addition, increased vagal input (hypervagotonia) has been proposed to constitute the direct cause of SND manifestations (122). Clinical studies have established a correlation between endurance athletic training and an increased risk of bradyarrhythmia, atrioventricular block (123), and atrial fibrillation (124) requiring PPM (125). Both hypervagotonia and intrinsic remodeling of SAN ion channel expression have been proposed as mechanisms of SND in endurance athletes. I_f and Hcn4 are downregulated in animal models of athletic training (126, 127). In addition, the heart rate of endurance athletes shows reduced sensitivity to the I_f blocker ivabradine, suggesting reduced HCN expression (126). Thus, it is possible that both changes in the sympathovagal balance and intrinsic remodeling of ion channels in SAN pacemaker cells contribute to bradyarrhythmia induced by athletic training.

SND secondary to cardiovascular disease is often associated with atrial tachyarrhythmia and atrial fibrillation. In early studies, SND was present in 5% of patients affected by atrial fibrillation for less than 1 year. However, SND incidence increased to 45% in patients with atrial fibrillation for 10 years or longer (38). In a canine model of atrial fibrillation induced by atrial tachypacing, SND was explained by a reduction in the expression of I_f channels (128). This mechanism can account for SND associated with chronic or persistent atrial fibrillation. However, because a significant number of patients present progressively worsening SND, in which association with atrial fibrillation and conduction dysfunction constitutes late comorbidities, it is an attractive hypothesis that early pharmacologic treatment of SND may prevent associated arrhythmias. Ischemic disease can also lead to SND. Typical examples of ischemic conditions favoring SND are acute stenosis or thrombosis of the SAN artery (129) and myocardial infarction (130). As discussed above, heart failure is a major provider of secondary SND. Several mechanistic aspects of how myocardial heart failure leads to SND have yet to be identified. However, in rabbit (131), canine (132), and mouse (133) models of heart failure, reduction of pacemaker activity has been attributed to downregulation of I_f and expression of its predominant channel isoform HCN4, with consequent SND.

Myocardial infarction can often degenerate to heart failure and consequently bring secondary SND forms. A study combining myocardial infarction with diabetic condition showed an increase in oxidized CaMKII, apoptosis of pacemaker cells, SND, and mortality in wild-type mice. These effects were prevented in knockin mice expressing oxidation-resistant CaMKII, in which paired methionines in the CaMKII regulatory domain mutated to valines (134). Oxidation of CaMKII could thus constitute an important mechanism in SND secondary to myocardial infarction and neurohumoral dysregulation, potentially leading to heart failure (121).

To date, several mechanisms have been proposed to account for age-related SND. (For a detailed discussion about age-related SND, see 135.) Similar to other secondary forms of SND, decline of pacemaker activity with aging is attributed to intrinsic remodeling of the SAN structure and expression of ion channels involved in automaticity. A decrease in intrinsic electrical coupling due to progressive tissue fibrosis has been proposed as a primary factor in age-related SND (see 136 for a review). However, some individuals with a high degree of SAN fibrosis can be under normal sinoatrial rhythm (137). In addition, work by Larson et al. in 2013 (138) demonstrated slowed intrinsic pacemaker activity and reduced densities of I_f, I_CaL, and I_CaT in SAN pacemaker cells of bradycardic aged mice. It is thus possible that both structural and ionic factors contribute to age-related SND in a patient-dependent manner.

PHARMACOLOGIC APPROACHES TO SINOATRIAL NODE DYSFUNCTION IN CURRENT CLINICAL PRACTICE

There are many potentially usable drugs for acute bradycardia and SND; however, the most widely used drugs are catecholaminergic agonists like isoproterenol, the muscarinic receptor inhibitor atropine, and the adenosine receptor blockers aminophylline and theophylline. Isoproterenol is a β-selective agonist devoid of vasoconstriction effects. Isoproterenol has shown some positive effects, ameliorating the heart rate in patients with bradycardia. However, administration of isoproterenol can also induce supraventricular tachycardia or β-receptor-dependent vasodilatation (139); its use is recommended only for intrahospital management of acute SND or electrophysiological evaluation of SND (4). Other catecholaminergic agonists such as dopamine or epinephrine present more complex effects because of their mixed α- and β-receptor activation (140). High doses of dopamine may be required to obtain a positive chronotropic effect. However, at these doses, the increase in heart rate augments the myocardial oxygen demand. Proarrhythmic effects have also been reported (4). Because increased oxygen consumption can be accompanied by coronary vasoconstriction, dopamine is not indicated in SND associated with cardiac ischemic disease (4). Epinephrine is a potent α- and β-receptor activator. At clinical doses epinephrine increases heart rate, blood pressure, and myocardial oxygen consumption. Consequently, use of catecholamines is indicated only under intrahospital and hemodynamic monitoring conditions (4). Atropine is a well-known inhibitor of muscarinic receptors. Clinical studies have showed an improvement in heart rate in patients with acute bradycardia and SND, including SND secondary to myocardial infarction (141, 142). Atropine is also used to diagnostically evaluate SND (143, 144). However, some adverse effects of atropine, including tachycardia and psychotic states, have been reported (141, 144). Theophylline and aminophylline belong to the pharmaceutical class of methylxanthines. Theophylline, in particular, is probably the most widely used drug to treat SND outside hospital settings. The beneficial effect of methylxanthines on heart rate is attributed to their ability to block adenosine receptors (145), which could make these drugs suitable for treating bradycardia in posttransplantation hearts (146, 147) and following spinal cord injury (148). In a limited clinical study, theophylline also improved heart rate, without significant adverse effects, which prevented the need for PPM implantation (149). Another promising drug is the phosphodiesterase inhibitor cilostazol, which is used to prevent thrombosis (150). In particular, some studies indicate that cilostazol improves the heart rate in patients with SND with tachycardia-bradycardia syndrome (151, 152). In addition, one study suggests that cilostazol can delay PPM implantation in patients with SND (153). Several mechanisms may help improve the heart rate in patients with SND treated with cilostazol. These include augmented sympathetic input induced by systemic vasodilatation and elevation of cAMP concentration in SAN pacemaker cells.

Drug intoxication constitutes an additional challenge for pharmacologic treatment of bradycardia attributable to SND (4). Frontline treatment of SND following intoxication with β-blockers, Ca²⁺ channels blockers, and digitalis (digoxin) exists. Intoxication with β-blockers can be treated by administering glucagon, which stimulates hepatic adenylate cyclase and thus leads to increased glycolysis. No systematic clinical studies that include cohorts of patients are available to validate the use of glucagon for SND secondary to intoxication with β-blockers (154). However, available data support the concept of antagonizing the effect of β-blockers and Ca²⁺ channels blockers with glucagon or insulin to improve the heart rate in patients who present symptoms of SND (155). Finally, intravenous bolus of calcium-gluconate can be used to counter the effects of Ca²⁺ channels blockers such as verapamil or amlodipine (4, 155). Digoxin intoxication induces complex arrhythmia patterns and bradyarrhythmia (156). Digoxin is a poorly dialyzable drug. Consequently, treatment of SND secondary to glycoside intoxication is applied using specific antidigoxin antibodies (157). Immunotherapy against digoxin intoxication is effective at improving the heart rate, with relatively low associated mortality (158).

NEW PHARMACOLOGIC TARGETS FOR SINOATRIAL NODE DYSFUNCTION IN ANIMAL MODELS

In the past decade, researchers have tested new potential pharmacologic approaches to treat SND by targeting GIRK channels. Our group has tested this approach by crossing Girk4^−/− mice with mice in which I_f conductance has been genetically silenced (25). Genetic ablation of I_KACh effectively rescued SND and atrioventricular dysfunction in these double-mutant mice and also prevented SND-associated ventricular arrhythmia (25). Isoproterenol was effective at elevating the heart rate of these mice but failed to correct SND hallmarks such as sinoatrial pauses and atrioventricular dysfunction (25), suggesting that direct inhibition of I_KACh could be more effective than catecholaminergic stimulation. Furthermore, genetic ablation of I_KACh was effective at rescuing SND and atrioventricular dysfunction in SANDD Ca_v1.3^−/− mice (26, 159). Tertiapin-Q is a synthetic stabilized form of tertiapin, a 21-residue peptide purified from honey bee venom (160, 161). Tertiapin-Q is a potent inhibitor of GIRK channels (162, 163). Previous work has shown that tertiapin-Q dampens the negative chronotropic and dromotropic effects of acetylcholine in isolated rabbit and guinea pig hearts (164). We thus hypothesized that pharmacologic inhibition of I_KACh could mimic the effects of genetic ablation of Girk4. Consistent with this hypothesis, we observed rescuing of SND after tertiapin-Q was administered to Ca_v1.3^−/− mice (26) (Figure 3). These studies suggest that loss of function in an ion channel contributing to pacemaking could be compensated for by interfering with I_KACh activation. This phenomenon of compensatory genetic or pharmacologic targeting of I_KACh can provide a new therapeutic option for SND in the future (26, 165) (see also Supplemental Appendix 3).

Figure 3.

Rescuing SND and atrioventricular dysfunction by targeting I_KACh. Tertiapin-Q rescues SND in Ca_v1.3^−/− mice. (Left) Typical tracing of ECG in Ca_v1.3^−/− mice that present with sinoatrial bradycardia and atrioventricular block. (Center) Close-up view of a structural model of tertiapin bound to the GIRK pore. K17, K21, and D164 are residues important for peptide activity. (Right) Administration of the GIRK pore blocker tertiapin-Q (tertiapin, central panel) normalizes the heart rate in Ca_v1.3^−/− mice. Figure adapted with permission from Reference 35. Abbreviations: ECG, electrocardiogram; GIRK, G protein–activated K⁺ channel; SND, sinoatrial node dysfunction.

Although no primary forms of SND linked to K_Ca channels have been described to date, evidence indicates that the activity of these channels can contribute to arrhythmia, atrioventricular dysfunction, and SND. Global deletion of SK2 channels in mice is proarrhythmic and induces atrioventricular block (75, 166), thus suggesting that altering the SK current could lead to SND. Consistent with this hypothesis, pharmacologically inhibiting the SK current recovers tachycardia-bradycardia syndrome in a model of Ca²⁺ overload caused by NCX1 deletion (32). Similarly, selective inhibition of the IK isoform SK4 was sufficient to reduce delayed afterdepolarizations and arrhythmic Ca²⁺ transients in a CPVT model (31). Mechanistically, hyperactivation of SK or IK channels in these models could lead to the hyperpolarization of SAN cells, as shown in a model of neurons (167). In support of this hypothesis, selective inhibition of SK4 channels rescued SAN arrhythmia and SND, suggesting that the family of K_Ca channels could be an additional pharmacologic target to treat SND.

HUMAN SINOATRIAL NODE AS MODEL OF SINOATRIAL NODE DYSFUNCTION

The human SAN complex is a specialized and heterogeneous intramural three-dimensional (3D) structure with multiple intranodal pacemakers and several specialized sinoatrial conduction pathways (SACPs), which are responsible for transmitting electrical impulses to the right atrium (168, 169). The human SAN pacemaker–conduction complex relies heavily on a sophisticated machinery of multiple molecular pathways that communicate within the 3D structure to efficiently maintain a physiologically relevant heart rate (49, 66) (Figure 4). However, there is a paucity of studies addressing mechanisms that contribute to automaticity and intranodal conduction directly in the human SAN complex at the molecular, cellular, and tissue levels (29).

Figure 4.

Functional and structural imaging of nondiseased human SAN and human SAN with a history of SND. The columns from left to right depict the three-dimensional reconstruction of human SAN, SAN activation exciting atria (top, nondiseased SAN) or exit block (bottom, SAN with SND), and histology staining of the SAN complex, with magnified section of lateral middle SACP showing continuously coupled myocytes (top inset) or fibrotic disruptions (bottom inset). Region of intranodal conduction block coincided with strands of intranodal fibrosis and the SAN artery. Green asterisks indicate the exit point of SAN activation to the atrial myocardium and the earliest atrial activation. Abbreviations: CT, crista terminalis; Endo, endocardium; Epi, epicardium; IAS, interatrial septum; SACP, sinoatrial node conduction pathway; SAN, sinoatrial node; SVC, superior vena cava. Adapted with permission from Reference 49.

Disease-induced remodeling of many of the molecular components critical to SAN function can lead to SND in humans (27, 169, 170). However, most of these molecular components critical to SAN function have been studied only in animal models, in which the SAN may have significantly different functional and anatomical features compared with the human SAN, especially aged and/or diseased human SAN with SND. The development of optimal treatments for SND will require in-depth knowledge about the mechanisms involved in robust regulation of human SAN rhythm. However, substantial gaps exist in data on SAN functions obtained directly from the human SAN (170, 171), and considerable understanding is inferred from animal models, which may not reproduce the human clinical SND phenomena (169), or from clinical electrogram recordings that are restricted to the atrial surface (168, 172).

In the last five years, studies of the ex vivo human heart (49, 66, 173) directly address these limitations (Figure 4; Supplemental Figure 4). These studies (49, 174) provide a unique opportunity to reinvent the translational study of human cardiac disease by applying state-of-the-art intramural mapping techniques, consisting of near-infrared optical mapping (27, 29, 66, 174), 3D structural imaging (49, 173), and molecular mapping (59), to resolve mechanisms of human SAN function in normal and diseased hearts that are not possible to study in vivo. To establish the efficient use of explanted human hearts with intact SANs, researchers have developed 3D integrative approaches to investigate live human hearts (with and without arrhythmia including SND) under ex vivo physiological conditions. For example, the cardiac transplantation program at the Ohio State University Wexner Medical Center (https://wexnermedical.osu.edu/transplant/heart) provides cardioplegically arrested explanted chronic failing hearts. The Lifeline of Ohio (http://lifelineofohio.org/), an independent nonprofit organ procurement organization, provides cardioplegically arrested donor hearts with and without histories of cardiac diseases and comorbidities (e.g., hypertension, diabetes, or history of chronic smoke and alcohol consumption/abuse) (174).

These multidisciplinary integrated approaches have identified that SND may be orchestrated by (a) heterogeneous fibrotic structural remodeling in SACPs (49, 136, 173) and by compartment-specific molecular remodeling in (b) I_f pacemaker channels (HCN1/HCN2/HCN4) (59), (c) adenosine A1 receptor and I_KAdo channels (GIRK1/GIRK4) (27, 29, 174) (Supplemental Figure 4), or (d) nNa_v (nNa_v1.6) and cNa_v1.5 isoforms in human SAN (66). In 2020, optical mapping studies revealed that unlike cNa_v1.5, nNa_v may predominantly contribute to SAN intranodal conduction rather than to atrial conduction. By contrast, cNa_v1.5 plays important roles in both SAN pacemaking and conduction, especially during adenosine or atrial pacing challenges, to prevent intranodal conduction failure. Impairment of nNa_v can lead to SAN exit block, disorganized intranodal pacemakers, and SAN micro- and macroreentry. Furthermore, these functional observations are supported by higher expression of nNa_v (Na_v1.1 and Na_v1.6) and lower expression of cNa_v1.5 in human SAN pacemaker cells versus the surrounding atrial myocardium. Importantly, several nNa_v transcripts were vulnerable to cardiac remodeling associated with heart failure, cardiac hypertrophy, and modifying risk factors such as history of chronic alcohol consumption, which could promote a substrate for SAN arrhythmias (66).

From a clinical perspective, studies of the ex vivo human SAN (66) highlight limiting the use of drugs that may block nNa_v channels especially when vagal tone is high or in patients with heart failure (175) and atrial fibrillation (176) with high plasma levels of adenosine. Furthermore, these studies suggest that region-specific SAN disease remodeling, such as fibrotic infiltration (136), significantly contributes to the intrinsic region-specific SAN conduction abnormalities and arrhythmias (Figure 4). Disease-induced structural remodeling may exacerbate region-specific conduction abnormalities induced by both Na_v blockers (66) and adenosine through GIRK channels and predispose to SND (Supplemental Figure 4). As such, specific GIRK4 channel blockers (e.g., tertiapin-Q) can prevent SAN pacemaker arrest and exit block induced by adenosine and parasympathetic hyperactivity and eventually be used in patients with atrial fibrillation and SND (29).

In summary, integrative studies of the ex vivo human SAN revealed that the availability of multiple redundant structural compartments (intranodal pacemakers and SACPs) and molecular components in the human SAN are important backup mechanisms to robustly protect SAN conduction and pacemaking, and prevent rhythm failure and SND, in the context of multiple disease-induced impairments of conduction (29, 136, 173).

CONCLUSIONS AND PERSPECTIVES

The development of new pharmacologic and molecular strategies to treat chronic primary and secondary forms of SND has been complicated by a lack of knowledge about the mechanisms underlying this complex pathology. However, researchers have described in the last few years new and unexpected mechanisms of SND by using human genetics, animal models of SND, and ex vivo SAN from human hearts with a history of secondary SND. These investigations have indicated potentially innovative pharmacologic targets such as GIRK and SK channels for the management of SND. Furthermore, functional exploration of animal models of SND and human SAN may help redirect the application of antiarrhythmic drugs and create innovative therapies for concomitant control of SND and associated arrhythmias. We expect that additional new SND mechanisms will be unraveled in the coming years, with extensive functional genomics investigation of primary familial SND and development of new cellular and animal models of secondary forms of SND.