Sinus node dysfunction: current understanding and future directions

Abstract

The sinoatrial node (SAN) is the primary pacemaker of the heart. Normal SAN function is crucial in maintaining proper cardiac rhythm and contraction. Sinus node dysfunction (SND) is due to abnormalities within the SAN, which can affect the heartbeat frequency, regularity, and the propagation of electrical pulses through the cardiac conduction system. As a result, SND often increases the risk of cardiac arrhythmias. SND is most commonly seen as a disease of the elderly given the role of degenerative fibrosis as well as other age-dependent changes in its pathogenesis. Despite the prevalence of SND, current treatment is limited to pacemaker implantation, which is associated with substantial medical costs and complications. Emerging evidence has identified various genetic abnormalities that can cause SND, shedding light on the molecular underpinnings of SND. Identification of these molecular mechanisms and pathways implicated in the pathogenesis of SND is hoped to identify novel therapeutic targets for the development of more effective therapies for this disease. In this review article, we examine the anatomy of the SAN and the pathophysiology and epidemiology of SND. We then discuss in detail the most common genetic mutations correlated with SND and provide our perspectives on future research and therapeutic opportunities in this field.

INTRODUCTION

First discovered by Keith and Flack in 1907 (1), the sinoatrial node (SAN) is a key component of the native cardiac conduction system (CCS) and is the primary pacemaker of the heart. The normal function of the SAN is crucial in maintaining proper cardiac rhythm, which affects cardiac performance. Sinus node dysfunction (SND), also historically known as sick sinus syndrome, is due to abnormalities within the SAN, which can affect the heartbeat frequency, regularity, and the propagation of electrical pulses through the CCS. As a result, SND often increases the risk of cardiac arrhythmias. As the most common indication for pacemaker implantations (2, 3), SND affects individuals of all ages and is prevalent not only in the United States but also worldwide. Emerging evidence has also established various genetic abnormalities that can cause SND, which could help us understand the molecular underpinnings of this disease. In this review article, we first briefly explain the anatomy of the SAN and the pathophysiology of SND, followed by the epidemiology of SND. We then discuss in detail the most common genetic mutations correlated with SND and current therapeutic options. We conclude the review with our perspectives on future research and therapeutic opportunities in this field.

ANATOMY AND PHYSIOLOGY OF THE SAN

Anatomically, the SAN is located at the anterolateral aspect of the junction of the superior vena cava and the right atrium (RA), lodged immediately subepicardially within the sulcus terminalis (4) (Fig. 1). Grossly, the human SAN has a crescent-like morphology, with a superior head, middle body, and inferior tail portion, although subtle variations in morphology are commonly observed (8). In humans, the mean SAN size is 13.5 mm in length and 1.5 mm in width along its body (8). The SAN body is the most epicardial structure, ∼0.1–1 mm from the epicardial surface, while the SAN tail and head are seen to penetrate into the musculature of the terminal crest, diving closer toward the endocardial surface (8). The distance between the SAN and the endocardium ranges from 2.3 to 4.6 mm in the majority of subjects, with the largest separation seen on the caudal surface of the SAN body, making this location the most resistant to endocardial ablation (8).

Figure 1.

Gross anatomy of human and mouse sinoatrial node (SAN). A: lateral view of a human heart specimen showing location of the SAN. Reproduced from Hayes et al. (5). B: posterior view of a X-Gal-stained mouse heart showing the SA nodal area. Reproduced from Lee et al. (6). C: dissected human heart showing SAN area (dotted circle). Reproduced from Mitrofanova et al. (7). D: dissected SAN tissue from a mouse. Approximate sizes are shown. Ao, aorta; CCV, common carotid vein; CT, crista terminalis; IAS, interatrial septum; IVC, inferior vena cava; LA, left atrium; RCV, right carotid vein; RA, right atrium; RAA, right atrial appendage; SVC, superior vena cava. Reproduced images, as indicated, are used with permission.

The SAN contains the highest density of pacemaker cells, which are histologically and functionally distinct from the surrounding myocytes. Histologically, these pacemaker cells are clear, pale, and relatively empty of the microfilament cytoskeleton and sarcomeric components compared with the surrounding atrial tissue (9). While the nodal cells are contained in a matrix of predominately fibrous tissue, the SAN itself is not isolated from the surrounding atrial myocardium (4). The SAN exhibits several radiations up to 2 mm long along its margins, which then interdigitate with atrial myocardium in a smooth transition, with cells along the peripheral zone seen to be transitional both in structure and function (8, 10). From a developmental point of view, the SAN is derived from the sinus venosus, which expresses Shox2 (encoding short stature homeobox 2), a gene that represses the activation of several genes involved in contractile myocardium (9).

The pacemaker cells are characterized by their automaticity, arising from their spontaneous (phase 4) diastolic depolarization (10) (Fig. 2). The sinoatrial (SA) nodal cells exhibit the most rapid phase 4 depolarization and are thus the dominant pacemakers in a normal heart. Other secondary locations of pacemaker cells include the atrioventricular (AV) node and the ventricular conduction system, which can serve as a backup pacemaker in case of SAN failure. The automaticity of pacemaker cells is the result of a complex interplay between a number of ionic currents, including the inward funny current (I_f), T-type and L-type Ca²⁺ currents (I_Ca,T, I_Ca,L), outward delayed rectifier K⁺ current (I_K), and electrogenic Na⁺/Ca²⁺ exchanger (I_NCX) (10) (Fig. 2). The relative importance of these currents involved in pacemaking remains controversial (9, 10). At approximately −65 to −40 mV, a cationic current I_f is conducted through the hyperpolarization-activated cyclic nucleotide-gated (HCN) channels which depolarizes the semipermeable membrane. Once the threshold potential is reached, the voltage-gated T-type Ca²⁺ channel (TTCC; I_Ca,T) and the L-type Ca²⁺ channel (LTCC; I_Ca,L) are activated, which increases the membrane potential and produce the upstroke of the action potential (AP) in SAN pacemaker cells. Of the two LTCC isoforms in the SAN (Cav1.2 and Cav1.3), Cav1.3 in particular is thought to play a key role in diastolic depolarization. Membrane repolarization starts because of the activation of the rapid (I_Kr) and slow (I_Ks) outward rectifier K⁺ currents along with simultaneous inactivation of the I_Ca,L current, which reduces the membrane potential of pacemaker cells. This process is known as the “membrane clock.” On the other hand, the automaticity of the pacemaker cells is also influenced by subsarcolemmal Ca²⁺ induced Ca²⁺ release (CICR) and membrane-bound Na⁺/Ca²⁺ exchanger (NCX) activity, known as the Ca²⁺ clock (11). The Ca²⁺ entry during diastolic depolarization can activate ryanodine receptors (RyRs) and initiate Ca²⁺ release from the sarcoplasmic reticulum (SR), which in turn activates NCX and produces an inward current during diastolic depolarization (12). Spontaneous local SR Ca²⁺ release that is not triggered by I_Ca,L during diastolic depolarization has also been observed to occur independent of the CICR mechanism; the relative contribution of this phenomenon to the Ca²⁺ clock remains unclear however (13). Ca²⁺ stores in the SR are then replenished by the sarcoplasmic-endoplasmic reticulum Ca²⁺ ATPase (SERCA) channel on the SR membrane (9). Many of these channels and Ca²⁺ handling proteins can be phosphorylated by protein kinase A (PKA) or Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) upon β-adrenergic stimulation (14–16). The synergistic interplay between the two fundamentally different systems provides a robust range of heart rates (HR), with the membrane clock helping to provide a stable minimum HR under increased vagal tone while the Ca²⁺ clock gives a wide dynamic range to dramatically accelerate the HR when sympathetic tone increases (17).

Figure 2.

Electrophysiology of pacemaker cells. A: ionic current in each phase of an action potential in a pacemaker cell. B: interplay between membrane clock and Ca²⁺ clock. Casq2, calsequestrin-2; PLB, phospholamban; RyR2, ryanodine receptor type 2; SERCA, sarco(endo)plasmic reticulum Ca²⁺-ATPase; SR, sarcoplasmic reticulum. B was created with BioRender.com and published with permission.

In addition to the membrane and Ca²⁺ clock, there is an increasing recognition of a “third clock,” also known as the “mechanics clock,” referring to the phenomenon of SAN mechanosensitivity (18). SAN stretch has long been known to cause an acute increase in HR in most mammals, termed the Bainbridge reflex (19). Interestingly, HR in mice tends to decrease in response to sustained stretch, likely because of differences in the relative contribution of stretch-sensitive channels to the species-specific AP (20). Regardless of the directionality in the stretch-induced changes in HR, it has been recognized that the mechanics clock is a regulator of SAN automaticity (18). While it was initially thought that the extracardiac neurohumoral pathways control the stretch-induced changes in HR, this enhanced automaticity in response to stretch is observed to occur in isolated hearts and even isolated SAN pacemaker cells, suggesting that intrinsic mechanosensitive mechanisms could control the mechanics clock of the SAN (18, 20). Mechanoelectric coupling is an increasingly recognized component of the automaticity of the SAN and perhaps the most important regarding maintaining in vivo global cardiac function. The SAN stretch mechanism is thought to entrain the numerous SAN myocytes by normalizing aberrant electrical activity and promoting synchronized excitation across the entire SAN (18). The molecular mechanisms underlying the mechanics clock remain incompletely understood. Cation nonselective stretch-activated channels (SAC_NS), which have a reversal potential between −20 and 0 mV, are potential candidates for this mechanism although the molecular identity of these channels remains elusive. Several channels involved in the membrane and Ca²⁺ clocks have also been demonstrated to be mechanosensitive, such as the HCN (21), LTCC (22), delayed rectifier K⁺, and NCX channels (23), providing evidence for the interdependent nature of these three clocks in the dynamic modulation of SAN automaticity.

While anatomically the SAN exists as a discrete unit, the site of origin of electrical activity has been suggested to shift within the SAN itself with changes in HR (24). This shift in the leading pacemaker within the SAN is reflected as subtle alterations in P-wave morphology. Recently, optical mapping applied to ex vivo RA tissue isolated from rats and humans revealed the presence of two competing pacemaker sites localized to the superior and inferior regions of the SAN, which preferentially control fast and slow HR, respectively (24). Molecular analysis of these two regions shows unique but similar transcriptional profiles, suggesting inferior SAN dominance when the superior SAN is silent (24). While having multiple pacemaker sites within the SAN may be advantageous from an evolutionary perspective as this system provides a backup site in the case of failure of the primary pacemaker site, further studies are needed to verify these findings.

The majority of our understanding of the molecular mechanisms underlying SAN function has been extrapolated from animal models given their ease of manipulation and practicality for research because of the logistical barriers of obtaining sufficient human SAN specimens. While the overall SAN function bears many similarities among mammalian species, structural and electrophysiological differences unique to each species are important to bear in mind when extrapolating findings from animal models to humans (Table 1). For example, as previously mentioned, mice SAN tissue uniquely demonstrates a reduction in HR in response to stretch, whereas most other mammals (including the dog, cat, rabbit, zebrafish, and guinea pig) demonstrate an increase in HR similar to humans (18). Furthermore, the predominant autonomic tone is seen to differ among species, with rodents, monkeys, and pigs exhibiting a predominant sympathetic tone in comparison to humans and dogs, which exhibit a predominant vagal tone (26). While the molecular mechanisms for these species-specific differences in SAN function remain to be fully elucidated, differences in both density and kinetic behavior of key ion channels including HCN4, the voltage-dependent Na⁺ channel (Nav1.5), and SERCA are thought to play a significant role (35). The global SAN architecture also varies across species, with the human, dog, and pig SAN surrounded by a layer of atrial muscle between the SAN and endocardium to protect the region against high wall stress, while smaller mammals such as the rabbit and guinea pig lack this layer (36). In addition, the extent of connective tissue within the SAN and the organization of myofilaments in nodal cells are highly variable across species (27). Taken together, these molecular and architectural differences in SAN structure among species are important to consider when extrapolating results from animal models.

Table 1.

SAN characteristics of select animal models compared with humans

Species	SAN Length, mm	SAN Width, mm	Normal HR, Beats/Min (25)	Cycle Length, ms (26)	Sinoatrial Conduction Time, ms (26)	Action Potential Amplitude, mV (26)	Action Potential Duration, ms (26)	Maximum Diastolic Potential, mV (26)	Diastolic Depolarization Rate, mV/s (26)
Human	8−21.5 (8)	1−5 (8, 27)	60−100	828 (28)	82 (29)	72	309	−62	48.9 (28)
Mouse	1.5 (30)		450−750	144	4	56	82	−51	117
Rabbit	6−7 (27)	2 (27)	180−350	386	32	66	180	−61	83
Dog	5−15 (27, 31)	2 (27, 31)	70−120	542 (31)	45 (31)	56 (32)	220 (33)	−56 (32)	72 (34)
Cat	7.4 (27)	2.2 (27)	120−140	418	29	60	202	−59	71
Guinea pig	4 (27)	1 (27)	200−300	275	12	67	137	−57	121

Displayed are data values (reference citation). SAN, sinoatrial node; HR, heart rate.

EPIDEMIOLOGY OF SND

Normal sinus rhythm is initiated by electrical impulses arising from the SAN. In normal sinus rhythm, the P wave is less than 120 ms in duration and 0.25 mV or less in height with a constant P-P interval on ECG. SND encompasses several related conditions that result from abnormalities in the intrinsic SAN impulse formation and/or propagation, leading to inappropriately low HRs, pauses, or irregularities in the heart rhythm (Fig. 3) (2). SND is typically diagnosed based on ECG recording. Electrophysiology studies are not routinely performed to diagnose SND, but they can be used in equivocal cases. The corrected sinus node recovery time (cSNRT) and sinoatrial conduction time (SACT) in response to atrial pacing are two common parameters that can aid the investigation of SND. Regardless of the underlying cause, SND is associated with increased cSNRT, conduction delay along the CCS, a unicentric and caudal shift of the leading pacemaker activity within the SAN, and increased areas of low voltage within the RA due to either atrophy or scar formation (9).

Figure 3.

Classifications of sinus node dysfunction. SAN, sinoatrial node.

Given the role of degenerative fibrosis and other age-related remodeling in the pathogenesis of SND, this disease is most commonly seen as a disease of the elderly and affects one out of 600 cardiac patients above the age of 65 (37). In a study by Jensen et al. (38), among 20,572 patients in the Atherosclerosis Risk In Communities (ARIC) and Cardiovascular Health Study (CHS) cohorts, 291 incident cases of SND (1.4% of the cohort) were reported over an average follow-up time of 17 yr, giving an unadjusted rate of 0.8 per 1,000 person-years. The incidence of SND is seen to steadily increase with age, with the greatest risk in patients in their 70–80s (2, 38). In the randomized Mode Selection Trial in Sinus-Node Dysfunction (MOST) trial, which evaluated the optimal pacemaker device in 2010 patients with SND, the median age was 74 yr old (39). Given the aging population in the United States, the annual number of new SND cases in the United States is projected to increase from 78,000 in 2012 to 172,000 in 2060 (38).

SND is the most common indication for pacemaker implantation, accounting for just over 50% of pacemaker implants in the United States (2, 3). The natural course of untreated SND is highly variable (40). Recurrent syncope is common in patients with previous syncope related to their bradycardia (41). While SND can be highly symptomatic and associated with decreased quality of life, isolated SND in itself does not appear to affect overall survival whether treated or untreated and the incidence of sudden cardiac death is extremely low (40, 42). However, SND can frequently coexist with structural heart disease and heart failure (HF), which also increase in prevalence with age. The increase in comorbidities with age may explain the poorer prognosis observed in some studies of patients with SND (42).

Risk factors associated with SND include greater body mass index, height, hypertension, right bundle branch block, and longer QRS duration (38). AV conduction disturbances can occur in a portion of patients with SND given the similar degenerative process underlying both conditions. The annual incidence of high-grade AV block following pacemaker implantation for SND is around 1.8% (43, 44). This risk is highest among patients with evidence of conduction disturbance (complete bundle branch block or bifascicular block), with an incidence of up to 35% within 5 yr of pacemaker implantation (43). However, the overall low incidence of high-degree AV block in most patients with SND suggests that the degenerative process of the specialized conduction system is slow and does not dominate the clinical course of disease (40).

SND can often coexist with atrial arrhythmias, as the same degenerative changes underlying the development of SND are linked to the development of atrial arrhythmias (2). The combination of sinus bradycardia alternating with periods of atrial tachyarrhythmia is termed tachycardia-bradycardia syndrome. Long-standing atrial arrhythmia in itself leads to fibrotic changes and reduced percentage of nodal cells, predisposing toward SND (45). Up to one out of five patients with atrial fibrillation (AF) are affected by SND, highlighting the close relationship between the two conditions (46). AF and SND also demonstrate similar differences in incidence by race, with white patients having the highest incidence of AF and SND (38). A cross-sectional analysis of the Multi-Ethnic Study of Atherosclerosis (MESA) cohort of 6,814 Americans in 2000–2002 indicated a significantly lower risk of AF in African Americans, Hispanics, and Chinese compared with White Americans (47). These findings coincided with results from the previous study by Jensen et al. (38) that reported increased incidence of SND in white patients as compared with black patients. While the cause of this difference by race is unclear, the similar trends observed with both arrhythmias provide some evidence of correlation in the development of SND and AF.

MOLECULAR MECHANISMS OF SND

Idiopathic SND most often occurs in the elderly and is commonly due to age-related degenerative fibrosis of the SA nodal tissue and surrounding atrial myocardium (2). However, idiopathic SND can also occur in the absence of fibrosis (48). Age-related electrical remodeling and gene expression also have an increasingly recognized role in idiopathic SND, as animal studies have shown age-dependent changes in several ion channels including HCN1, HCN4, Nav1.5, and RyR2 associated with SND physiology (9, 49). In addition to age itself, frailty has been demonstrated to be an important determinant of SND in mouse models via electrophysiological and structural remodeling. Fibrosis, AP morphology, and changes in ion channel expression are correlated with frailty in aging mice (50, 51). Acute myocardial ischemia is another cause of SND and is due to injury to SAN tissue and/or abnormal autonomic tone (9). Although incidence of coronary artery disease and SND increase with age, whether chronic ischemia directly causes SND is unclear (9). Finally, rarer causes of SND include familial SND, inherited cardiomyopathies, or inherited arrhythmia syndromes associated with SND such as catecholaminergic polymorphic ventricular tachycardia (CPVT) and long QT syndrome (LQTS), which are often attributed to a specific gene mutation. While uncommon, such identified gene mutations provide tremendous insight into novel pathways and mechanisms involved in normal SAN function. Below we discuss the molecular mechanisms causing SND by impacting 1) the membrane clock, 2) the Ca²⁺ clock, or 3) other significant factors that do not strictly fall into these two categories.

Molecular Mechanisms Impacting The Membrane Clock

Hyperpolarization-activated cation currents.

Hyperpolarization-activated cation currents (I_f) were discovered in heart and nerve cells over 40 yr ago and are key determinants of pacemaker activity. The hyperpolarization-activated cyclic nucleotide-gated (HCN) gene family encodes the channels that underlie I_f (52). Among the four subtypes of HCN channels (HCN1–4), HCN1, 2, and 4 isoforms are expressed in SAN tissue. HCN4 is the predominant isoform expressed throughout the SAN while HCN1 is restricted to the head region of the SAN (53) and HCN2 shows a confined expression to individual pacemaker cells mainly in the periphery of the SAN (54). HCN channels can be activated upon membrane hyperpolarization with a variable threshold between −50 and −65 mV, leading to Na⁺ and K⁺ influx into the SAN pacemaker cells and generating cationic currents during the slow diastolic depolarization phase (55). Ivabradine, a drug used as a HR-reducing agent, has been shown to cause a current-dependent block on open HCN channels (9). When Ivabradine enters the pacemaker cells, it can bind to the open state of HCN channels, which can deactivate the channel during the depolarization phase. During hyperpolarization, the molecules are displaced by a current driving force, removing them from the pores. A HCN4 mutation (p.S672R) located within the cyclic nucleotide-binding domain (CNBD) of HCN4 was the first HCN mutation reported to be associated with a case of familial asymptomatic sinus bradycardia that was inherited in an autosomal-dominant pattern (56). The S672R mutation was seen to modify channel kinetics by shifting the current activation range to hyperpolarized voltages and slowing current deactivation (56). Since then, genetic mutations of HCN channels have been associated with a wide variety of cardiac phenotypes in addition to SND, including ventricular arrhythmias, left ventricular noncompaction, Brugada syndrome, and structural abnormalities of the myocardium (57), suggesting that the function of HCN channels can go beyond pacemaking (58).

Various HCN knockout mice models have been developed for the different HCN isotypes, each of which displays a unique cardiac phenotype. HCN4 plays a fundamental role during cardiac development, with constitutive global Hcn4 knockout mice demonstrating embryonic lethality (59). Isolated hearts with Hcn4 deletion demonstrate bradycardia with a 36.7% reduction in HR compared with controls, in addition to a 75–90% reduction in the I_f current (60). Inducible Hcn4 knockout mice interestingly demonstrate differing cardiac phenotypes based on the specific mutation and can generally be grouped into two phenotypes: recurrent sinus pauses with normal basal HR and normal response to isoproterenol versus pronounced bradycardia and AV block with decreased response to isoproterenol (61). The reasons for the differing phenotypes among the different Hcn4 knockout models remain unclear (60). Hcn2 knockout mice demonstrate sinus dysrhythmia at rest but normal basal and maximal HR as compared with controls in addition to a reduction of I_f current by ∼30% (62). Finally, Hcn1 and Hcn3 knockout mice do not appear to demonstrate any appreciable SND phenotype (61).

The activity of HCN channels is controlled by intracellular cyclic AMP (cAMP) levels. Increases in cAMP concentration within the physiological voltage range enhance HCN channel activity (63). A 1-bp deletion in exon 5 of HCN4 was detected in a patient with idiopathic SND, which produced a truncated HCN4 protein (hHCN4-573X) lacking the CNBD (64). Transgenic mice with cardiac-specific overexpression of the dominant-negative CNBD-deficient HCN4 subunit (hHCN4-573X) exhibit the loss of the cAMP sensitivity of I_f and lowered maximum firing rates of SAN pacemaker cells, although the overall range of SAN frequency regulation (difference between maximum and resting HR) is still preserved (65). On the other hand, heterozygous HCN4-R669Q knockin mice, where the cAMP-binding site of the CNBD was disabled, display normal HRs at rest and during exercise but exhibit SAN block and sinus pauses upon β-adrenergic stimulation (66). In contrast to the hHCN4-573X transgenic mice, the heterozygous HCN4-R669Q mice likely retain some cAMP sensitivity of channels, which may explain the different SND phenotypes observed between the two (65). Nevertheless, these results demonstrate that the cAMP-mediated regulation of I_f plays an indispensable role in HR adaptation during physical activity, and defects in this process promote SND.

Inward rectifier potassium 3 channel.

The inward rectifier potassium 3 (Kir3) channel is a tetrameric channel that more favorably conducts inward K⁺ current at hyperpolarized potentials (I_Kir) (67, 68). Among the four known subunits of the Kir3 family (Kir3.1–Kir3.4), Kir3.1, and Kir3.4 are expressed in the heart and encoded by KCNJ3 and KCNJ5, respectively (68). Kir3.1 and Kir3.4 constitute the heterotetrametric Kir3 channel at 1:1 stoichiometry in cardiomyocytes. The activation of Kir3.1/Kir3.4 channels are voltage independent and directly regulated by levels of membrane phosphatidylinositol 4,5 bisphosphate (PIP₂) (69). Additionally, because the Gβγ heterodimeric subunits play a crucial role in modulating the activity of Kir3.1/Kir3.4 channels upon vagal stimulation, Kir3.1 and Kir3.4 channels are also known as the G protein-activated inwardly rectifying K⁺ channel 1 and 4 (GIRK1, GIRK4) (70, 71). Activation of Kir3.1/Kir3.4 (or GIRK1/4) channels can lead to hyperpolarization of the membrane potential in SAN pacemaker cells, thus slowing HR (72). On the other hand, inactivation of Kir3.1 or Kir3.4 can induce sinus arrhythmias. Mutations of KCNJ5 have been associated with cardiac arrhythmias such as AF, LQTS, SND, and Andersen-Tawil syndrome with symptoms of periodic paralysis, cardiac arrhythmia, and dysmorphic heart features (73). The G→A nonsynonymous substitution at nucleotide 739 affects the composition of the Kir3.4 subunit and, as a result, impacts the activation of the cardiac GIRK channels (71, 74). A gain-of-function (GOF) mutation of KCNJ5 (p.W101C) identified in familial patients with SND is associated with early onset of sinus bradycardia, sinus arrest episodes accompanied with junctional ectopic heartbeats, and periods of AV block, specifically the Mobitz 2 type (71). GIRK4-deficient mice display shorter SNRT and sinus cycle length in the presence of carbachol (75). The mutated Kir3.4 protein can cause a loss of rectification by altering the stoichiometry of the Kir3.1 and Kir3.4 subunits in the channel and increasing K⁺ efflux and hyperpolarization of SAN pacemaker cells, thereby hampering depolarization and lowering HR (71).

Guanine nucleotide-binding protein.

Guanine nucleotide-binding protein (GNB) or G protein regulates the extracellular signaling pathways between the G protein-coupled receptors and effector proteins (76). As mentioned earlier, βγ-subunits of G proteins directly regulate the activation of GIRK1/4 channels in SAN pacemaker cells. Gehrmann et al. (77) have shown that muscarinic stimulation by carbachol can induce bradycardia and increase HR variability and baroreflex sensitivity in wild-type mice, which can be blunted by cardiac-specific overexpression of the nonprenylated Gγ2-subunits in mice. Among the different subtypes of Gβ subunit, Gβ2 and Gβ5 proteins, encoded by GNB2 and GNB5, respectively, are often involved in the regulation of GIRK1/4 channels or the acetylcholine-activated K⁺ current (I_K,ACh) (76). GOF mutations of GNB2 and GNB5 have been associated with SND (71, 76, 78, 79). Stallmeyer et al. (78) described a novel GNB2 mutation (p.R52L) associated with nonsyndromic SND in a three-generation family with autosomal dominant SND and AV block. The GOF mutation caused sustained activation of GIRK channels, leading to hyperpolarization of the cardiac membrane potential and reduced spontaneous activity (78). Later studies of human induced pluripotent stem cell differentiated cardiomyocytes (iPSC-CMs) lines with a recessive GNB5 mutation (p.S81L) displayed an augmented current density of I_K,ACh and a profound decrease of spontaneous activity as compared with isogenic controls (80). Interestingly, the application of the specific I_K,ACh blocker XEN-R0703 reversed the cellular phenotypes in the p.S81L iPSC-CMs, positioning it as a potential therapy in patients carrying GNB5 mutations. Additionally, inhibition of I_K,Ach by tertiapin-Q peptide can improve the HR and cardiac conduction in various genetic murine models of SND (81). These studies not only suggest the potential usage of the pharmacological I_KACh inhibitors for SND management but also further confirm the importance of keeping GIRK or I_K,Ach in check, thereby maintaining the repolarization of the membrane in SAN pacemaker cells.

Voltage-gated Ca²⁺ channels.

Both T-type and L-type voltage-gated Ca²⁺ channels (TTCCs, LTCCs) contribute to pacemaking activity in the heart. While the LTCCs are widely expressed throughout the myocardium and responsible for the initiation of cardiomyocyte contraction and pacemaker activity, TTCCs are primarily expressed in the CCS (12). The dihydropyridine (DHP)-sensitive LTCCs are constituted by the pore-forming α₁-subunit, and several accessory subunits and can be activated between −50 mV and −30 mV. Among the four α₁-subunits of LTCCs (Cav1.1–Ca1.4), Cav1.2 (encoded by CACNA1C) and Cav1.3 (encoded by CACNA1D) channels are involved in the automaticity of pacemaker cells. While both Cav1.2 and Cav1.3 channels respond to β-adrenergic stimulation similarly, the Cav1.3-based LTCCs can be activated at more negative voltages (−45 vs. −25 mV) with slower inactivation kinetics compared with the Cav1.2-based LTCC (82, 83). It is believed that Cav1.2 channels contribute to the upstroke phase of the AP and Cav1.3 channels contribute to diastolic depolarization in SAN pacemaker cells. When compared with LTCCs, the characteristics of TTCCs include a negative shift of voltage threshold for activation, a slower activation, and fast voltage-dependent inactivation (84). There are three α₁-subunits of TTCCs (Cav3.1–Cav3.3), with Cav3.1 (encoded by CACNA1G) being the most studied TTCC in SAN cells (12). The precise contribution of I_Ca,T to diastolic depolarization is not completely understood. However, emerging evidence suggests that the activation of TTCCs is involved in the local control of Ca²⁺ release within SAN pacemaker cells (85, 86). Genetic variants of CACNA1D have been commonly found in patients with sinoatrial node dysfunction and deafness (SANDD) (87, 88). Cacna1d^−/− mice exhibit SND and deafness, recapitulating the phenotype in patients with SANDD (89). Genetic variants of CACNA1G in humans have also been linked to bradycardia and congenital heart block (90).

Molecular Mechanisms Impacting the Ca²⁺ Clock

Na⁺/Ca²⁺ exchanger 1.

The Na⁺/Ca²⁺ exchanger 1 (NCX1) (encoded by SLC8A1) plays a critical role in maintaining Ca²⁺ homeostasis and is necessary for maintaining the automaticity of pacemaker cells (86, 91, 92). While at present, there have not been any NCX1 mutations identified in humans with SND, mice models have provided evidence for the critical role of NCX1 in normal SAN function. Mice with genetic inhibition of Slc8a1 (Slc8a1^−/−) exhibit a normal resting HR but a blunted response to the β-adrenergic stimulation-induced increase in HR (91). Interestingly, in the absence of fight-or-flight response of HR in Slc8a1^−/− mice, I_f plays an augmented role in β-adrenergic induced HR increases (91). In the study by Gao et al., Slc8a1^−/− mice were still seen to retain ∼20% of the control level NCX current, which may explain the normal basal HR observed (91). On the other hand, mice with complete atrial-specific knockout of NCX1 demonstrate a junctional escape rhythm and the complete absence of atrial depolarization (93). Despite the complete lack of atrial depolarization, isolated SAN cells from these knockout mice still demonstrate relatively normal I_f and intracellular Ca²⁺ stores, suggesting that the membrane clock (I_f) is not sufficient to spontaneously cause SAN depolarization in the absence of NCX1 (93). While spontaneous depolarization does not occur in these knockout mice, external pacing is still capable of causing the depolarization and conduction of atrial tissue, indicating intact electrical coupling between cells (93). Finally, NCX1 knockout mice exhibit loss of transverse axial tubules, invaginations of the surface sarcolemma in cardiomyocytes, presumably as a consequence of abnormal Ca²⁺ regulation and activation of Ca²⁺-dependent remodeling (94).

Ryanodine receptor type 2.

Ryanodine receptor type 2 (RyR2) protein (encoded by RYR2) can form a homotetramer channel located on the SR (95, 96). RyR2 channels mediate CICR from the SR, a critical step in the electrical contraction coupling process in working cardiomyocytes (97). Although the relative importance of RyR2 channels in maintaining baseline SAN automaticity may vary in different conditions, the involvement of RyR2 in regulating the automaticity of SAN pacemaker cells is well acknowledged. Conceptually, Ca²⁺ release via RyR2 channels can activate NCX current, potentially facilitating AP firing at the end of diastolic depolarization and supporting the Ca²⁺ clock. Experimentally, it has been shown that the classic RyR2 inhibitor ryanodine can reduce the firing rate of the SAN (92, 98). Moreover, it is known that PKA-mediated phosphorylation can enhance the activity of RyR2 channels. Because the basal levels of cAMP, an activator of PKA, are high in SAN pacemaker cells, the increased PKA-phosphorylation of RyR2 channels may cause more Ca²⁺ release and influence pacemaking activity in SAN pacemaker cells (14). Some evidence suggests that the Ca²⁺ clock and RyR2 may exert greater contribution to pacemaking activity at the periphery of SAN than at the center of the SAN, because of higher protein levels of NCX1, RyR2, and SERCA in the periphery of the SAN (99). There is ample evidence that dysfunctional RyR2 channels, typically with enhanced activity, can cause HF, AF, ventricular tachycardia, and other arrhythmias (100, 101). Genetic variants of RYR2 are often found in patients with CPVT and arrhythmogenic right ventricular cardiomyopathy/dysplasia ((102, 103). The direct association between the genetic variants of RYR2 and idiopathic SND is currently lacking. However, some CPVT patients with RYR2 mutations develop concomitant SND (104, 105). The CPVT-mimicking RyR2-R4496C heterozygous knockin mice display sinus pause following isoproterenol injection. The frequency of Ca²⁺ transients in isolated SAN cells as an indicator of pacemaking activity is comparable at baseline between wild-type and RyR2-R4496C mutant mice (106, 107). However, there is a blunted chronotropic response to β-adrenergic stimulation and increased frequency of sinus pauses in SAN cells of RyR2-R4496C mice, which are associated with the reduced I_Ca,L density (106). Mechanistically, it is speculated that the increased Ca²⁺ release via RyR2 channels can cause a rise in the cytosolic Ca²⁺ concentration and subsequently inactivate I_Ca,L, which can lead to impaired diastolic depolarization in SAN pacemaker cells.

Calsequestrin-2.

Calsequestrin-2 (Casq2) (encoded by CASQ2) is a SR Ca²⁺ binding protein and plays a key role in Ca²⁺ homeostasis by acting as a Ca²⁺ reservoir within the SR of cardiomyocytes. CASQ2 loss-of-function (LOF) mutations are associated with an autosomal recessive form of CPVT and represent the 2nd most frequent cause of CPVT after RYR2 mutations (108). Patients with CPVT typically display concomitant sinus bradycardia, which is often the only resting ECG abnormality (104, 109). Casq2 homozygous knockout mice (Casq2^−/−) exhibit atrial arrhythmias including atrial ectopic activity and bradycardia under sympathetic stimulation (110). The SAN conduction blocks in Casq2^−/− mice are associated with abnormal Ca²⁺ release and increased interstitial fibrosis within the atrial pacemaker complex, which may disrupt SAN pacemaking and enhance latent pacemaker activity.

Ankyrin-B.

Ankyrins, considered multifunctional proteins, are required to target ion channels and transporters in diverse tissues and cells (111). Ankyrin dysfunction has been linked to abnormalities of ion channels and is implicated in several arrhythmias such as LQTS, HR variability, CPVT, conduction defects, AF, bradycardia, and arrhythmogenic cardiomyopathy (112, 113). Ankyrin-B (AnkB) protein, encoded by Ankyrin-2 (ANK2), is an adaptor and scaffold protein that targets NCX1 and Na⁺/K⁺-ATPase in cardiomyocytes (112). Mutations of ANK2 were identified in patients with LQTS with concurrent SND (114, 115) and patients with familial SND and AF (116). In the latter study, severe and highly penetrant SND in two families were linked to LOF mutations in ANK2 (p.E1425G), the first nonion channel protein to be implicated in human SND (116). Individuals with the ANK2 mutation exhibited SND at all ages, including in utero. The Ank2 haploinsufficient mouse model produces a similar SND phenotype (114, 116, 117). Mechanistically, AnkB deficiency can cause a reduction in NCX and Na⁺/K⁺-ATPase protein levels, and subsequently lead to an increase in total cellular Ca²⁺, thereby impacting the Ca²⁺ clock (117).

Mitochondria-SR communication.

It is known that the high basal levels of cAMP in SAN cells are associated with a high mitochondrial respiratory rate (15, 16). Emerging evidence reveals that proper communication between mitochondria and the SR is important in modulating the Ca²⁺ and cAMP signaling in SAN cells. Disruption at the mitochondria-SR contact sites due to the reduced level of mitofusin-2 impairs Ca²⁺ release and cAMP signaling at the mitochondria-SR microdomains, leading to reduced spontaneous firing rate and prolonged AP duration in SAN cells (118). This mechanism contributes to HF-associated SND (118).

Sarcoplasmic-endoplasmic reticulum Ca²⁺ ATPase.

The sarcoplasmic-endoplasmic reticulum Ca²⁺ ATPase (SERCA) channel, embedded on the SR surface, serves to transport cytosolic Ca²⁺ back into the SR and hence plays a key role in cellular Ca²⁺ homeostasis. cAMP-activated PKA has been shown to phosphorylate SERCA, thereby modulating Ca²⁺ cycling in SAN cells (119). Inhibition of SERCA with cyclopiazonic acid leads to a dose-dependent suppression in spontaneous SAN firing by up to 50% (120). SERCA2a (the cardiac isoform) knockout mice exhibit normal HR but prolonged PR interval at baseline, as well as a reduced maximal HR during exercise (121). SERCA2a is known to be downregulated in HF, leading to aberrant Ca²⁺ cycling and a decreased force-frequency response (122). The abnormal expression of SERCA2a underlying both HF and SND may offer another mechanistic link between the two comorbidities.

Other Factors

Voltage-gated sodium channel.

Voltage-gated sodium channel type 5 alpha subunit (SCN5A) is highly expressed in cardiomyocytes and encodes Nav1.5 protein. Nav1.5 channels are responsible for the initiation and transmission of APs in the myocardium and determine the electrical conduction through the heart (123). GOF mutations in SCN5A underlie the development of LQTS because of the increased Na⁺ influx into cardiomyocytes, while the LOF mutations in SCN5A are often associated with Brugada syndrome and dilated cardiomyopathy (DCM) (123, 124). Mutations of SCN5A have been identified in patients with congenital SND (125, 126). Scn5a heterozygous knockout (Scn5a^+/−) mice also display slower SA conduction and overall reduced HR compared with wild-type littermates (127). Nav1.5 is likely absent in the central SAN pacemaker cells but is detectable in peripheral SAN myocytes. The impaired function of the Nav1.5 channels does not directly alter the spontaneous firing of the central SAN pacemaker cells, as the latter is dependent on the I_f-mediated membrane clock and the NCX-mediated Ca²⁺ clock as discussed above. However, in Scn5a^+/− mice, the peripheral SAN myocytes demonstrate a slowed pacemaker rate as compared with the central SAN myocytes, which contributes to the reduced HR in Scn5a^+/− mice. Additionally, the failure of impulses to conduct into adjacent atrial myocardium due to reduced excitability of the cardiomyocytes is likely the mechanism underlying the SCN5A-associated SND (127).

Myosin heavy chain 6.

Cardiac alpha-myosin heavy chain (encoded by MYH6) is an important sarcomeric protein that plays a significant role in cardiac contraction (128). MYH6 is dynamically expressed in cardiomyocytes and is essential for heart development. MYH6 homozygous mice exhibit embryonic lethality at embryonic days 11–12 with gross heart defects and survived heterozygotes mice display severe impairment of cardiac contraction and relaxation due to fibrotic lesions and sarcomeric structure alterations (129). Knockdown of MYH6 can lead to abnormal development of the atrial septum with immature actin filaments and abnormal cardiac looping (130, 131). Genetic variants of MYH6 are a well-known cause of hypertrophic cardiomyopathy and DCM, as well as congenital heart diseases (132, 133). MYH6 variants have also been found in patients with early onset and familial SND, AF, ventricular arrhythmias, and cardiac arrest (134–137). However, the molecular mechanism of MYH6 in SND and other cardiac arrhythmias remains to be elucidated. The available mechanistic studies have suggested that microRNA (miR)-208, a highly conserved cardiac-specific miR encoded by intron 29 of MYH6, is required for proper cardiac conduction. miR-208 can modulate the expression of cardiac gap junction protein connexin 40 (Cx40), and altered miR-208 levels can affect SAN impulse formation and atrial propagation (138, 139).

Short-stature homeobox 2.

Short-stature homeobox 2 Shox2 protein (encoded by SHOX2) is a homeodomain transcription factor, essential for the proper formation and development of the SAN. Several transcription factors, including Tbx18, Tbx3, Isl1, Tbx5, Bmp4, Nkx2.5, and Pitx2c are required for SAN development and function (140–146). Shox2 interacts with these genes to form a fine regulatory network controlling SAN morphogenesis through increasing the expression of Tbx3, Bmp4, and Islet1, being promoted by upstream Tbx5 and repressed by Nkx2.5 and Pitx2c (140). Homozygous deletion of Shox2 in mice leads to embryonic lethality due to severe cardiovascular defects, including bradycardia, severe hypoplastic SAN, and sinus valves owing to a significantly decreased level of cell proliferation (147, 148). The Shox2-deletion can result in reduced expression of HCN4 and Tbx3 (another key regulator of the CCS), along with ectopic activation of the chamber differentiation marker Nppa and atrial myocardium-specific gap junction protein Cx40 in the SAN region (147). On the other hand, overexpression of SHOX2 enhances the differentiation of SAN cells (149). Mutations of SHOX2 are mainly associated with the short stature phenotype seen in patients with Turner’s syndrome (150, 151). A few cases of SHOX2 genetic variants have been found in patients with early onset and familial forms of AF and SND (152–154). However, mechanistic studies are needed to understand the pathogenic mechanism of Shox2-related heart disease.

Lamin A/C.

The LMNA gene encodes A-type nuclear lamins, including lamin A and C (lamin A/C) (155). Lamin A/C proteins constitute the nuclear lamina and play an important role in cell differentiation, lineage specification, and tissue development (156). Nuclear lamin A/C interact with cytoskeleton components in the cytoplasm through the linker of nucleoskeleton and cytoskeleton (LINC) complex and chromatin in the nucleoplasm. These interactions support their function in mechanotransduction, nuclear stability, chromatin organization, gene regulation, DNA replication, cell-cycle regulation, and RNA splicing (157–161). Genomic regions that are bound to nuclear lamins are referred as lamina-associated domains (LADs), which are predominantly located at the nuclear periphery where the repressive chromatin marks are enriched (162, 163). Through LADs, lamin A/C participate in genomic organization (164), recruitment of epigenetic regulators (165), and gene expression (166). Genetic variants of LMNA are associated with a wide spectrum of diseases including DCM, conduction disorders, Hutchinson-Gilford Progeria Syndrome, Emery-Dreifuss muscular dystrophy, and restrictive dermopathy, which are collectively referred to as laminopathies (167). LMNA mutations are the second most common variants associated with familial cardiomyopathy, accounting for ∼6–8% of idiopathic DCM cases and 33% of familial DCM cases with conduction disorders (168–170). Additionally, LMNA mutations have been increasingly described in patients with SND, sinus arrest, AV conduction block, his/bundle disease, AF, supraventricular tachycardia, and ventricular tachycardia/fibrillation (168, 171). Lamin A/C knockout (Lmna^−/−) mice demonstrate progressive cardiac dilatation, myocardial fibrosis, and reduced HR with lengthened PR and QRS intervals, leading to premature death by 8 wk of age (172, 173). However, the molecular and cellular mechanisms underlying LMNA mutation-mediated pathogenesis of SND and cardiac conduction disorders remain unknown. Because many genes and pathways are reportedly dysregulated in diverse cell types including cardiomyocytes and cardiac fibroblasts due to lamin A/C deficiency (171, 173–175), it is likely that the pathogenic mechanisms contributing to the LMNA-associated SND and cardiac conduction disorders are multifaceted.

Source-sink mismatch.

For normal SAN function, a delicate source-sink balance must be maintained between the relatively few pacemaker cells (source) and surrounding atrial myocardium (sink) for the SAN to drive the larger surrounding myocardium without being overcome by myocardium’s hyperpolarizing influence (176). Altered connexin function and fibrotic remodeling can often cause a source-sink mismatch, thereby impairing the effectiveness of SAN function.

Connexins form gap junction channels which are responsible for proper AP propagation throughout the heart. Differential expression of connexins within the SAN architecture help maintain the source-sink balance by electrically uncoupling the SAN from the surrounding myocardium except at limited exit sites, reducing the perceived electrical load on the SAN (176). Connexin (Cx)43, the principal connexin of working myocardium, is absent from the center of the SAN and is only present in the periphery of the SAN where strong electrical coupling is needed to drive the atrial muscle (177). The SAN is further uncoupled from the surrounding tissue by the preferential expression of Cx45, which is characterized by its low conductance and is the major isoform expressed in the SAN (176). Altered expression of connexins has been associated with SND. A Cx45 mutation (p.R75H) in the GJC1 gene was described in two unrelated families (a de novo French case and a 3-generation Japanese family) who presented with atrial standstill and congenital AV block in addition to other extracardiac abnormalities including finger deformities, dental dysplasia, and craniofacial abnormalities (178). GJC1 conditional knockout mice also exhibit intermittent atrial pauses and SND but without AV node dysfunction, likely due to species differences in connexin expression in the AV node (178). Age-related decreased expression of Cx43 has also been observed in studies of guinea pigs, with conduction velocity subsequently decreasing with age (179). These age-related changes in connexin expression are thought to contribute to a sink-source mismatch between the SAN and atrial myocardium, providing another potential mechanism for age-related SND.

As mentioned earlier, pacemaker cells are enmeshed in fibrous connective tissue, which provides a structural scaffold and electrical insulation for the pacemaker cells (180). The fibrotic content within the SAN region increases with heart size and age in various species (181–184). Fibroblasts also express stretch-activated channels and can be electronically coupled with SAN myocytes; therefore, the healthy number of fibroblasts within the SAN can meet the contractile demands during the natural growth and prevent the SAN from overstretching (20, 183). However, excessive fibrosis is associated with a spectrum of SND including bradycardia, sinus pause, exit block, tachy-brady syndrome, and reentrant arrhythmia within the SAN. Increased fibrosis can interrupt the coupling between fibroblasts and pacemaker cells, thereby promoting a source-sink mismatch (185). Although mechanistic studies that directly address fibrotic remodeling within the SAN have been limited, it is generally thought that the enhanced crosstalk between pacemaker cells and fibroblasts driven by transforming growth factor (TGF)-β1 signaling can promote myofibroblast transdifferentiation and increase fibrosis in the SAN region (186). A recent transcriptome and proteomics study has shed a light on the fibroblast-specific expression profiles of human SAN in HF hearts compared with non-HF hearts (31). Specifically, the CILP1 (cartilage intermediate layler protein 1)- or POSTN (periostin)-positive fibroblasts were enriched in the SANs of HF hearts but not in non-HF hearts. Moreover, a recent study has shown that mice with CCS-specific deletion of Hippo kinases develop SND, which is associated with increased SAN fibrosis. Chromatin sequencing further revealed that the inactivation of Hippo kinases increases fibroblast proliferation within the SAN region by upregulating Tgfb1 (encoding TGF-β1) and Tgfb3 (encoding TGF-β3) in pacemaker cells (187). Interestingly, the genes involved in cell-cell adhesions, such as ZO-1 and Cdh2, are also altered in the SANs of CCS-specific Hippo knockout mice, suggesting a source-sink mismatch induced by the Hippo inactivation (187).

MANAGEMENT OF SND

In patients symptomatic with SND, assessment of potentially reversible or treatable causes of SND is the first step in management. These include acute myocardial infarction, atrial tachyarrhythmia, electrolyte or metabolic disturbances, hypothyroidism, or medications (2). However, most causes of SND tend to be chronic and irreversible (2). In these cases, there is currently no known cure, and the treatments that do exist for SND, for the most part, aim to manage symptoms from bradycardia attributable to SND. Outlined below are some of the main SND management methods:

Permanent Pacemakers

Permanent pacemaker implantation is the primary therapeutic intervention for patients symptomatic with SND without an identifiable reversible cause (2). Either single-chamber atrial pacemakers or dual-chamber pacemakers are used, depending on whether AV conduction abnormalities are present (Fig. 4). While single-chamber ventricular pacing is also feasible for SND, this method is a less physiological form of pacing due to loss of AV synchrony and can lead to chronic left ventricular dysfunction (188). Pacemaker syndrome, or development of signs and symptoms of HF or hypotension, is common with ventricular pacing, occurring in up to 20% of patients (39). While previous randomized controlled trials have failed to demonstrate a clear reduction in hard outcomes of death, stroke, or HF hospitalization with atrial-based pacing modes which preserve AV synchrony compared with ventricular pacing in SND (2), atrial pacing shows a significantly decreased risk of subsequent AF and mild improvement in quality of life (39, 189). Thus current guidelines recommend atrial pacing modes over single-chamber ventricular pacing for SND, primarily because of their decreased risk of subsequent AF (2).

Figure 4.

Dual-chamber pacemaker. Standard dual-chamber pacemaker consisting of a pulse generator in the upper chest and leads implanted into the right atrium and right ventricle.

In patients with SND and intact AV conduction, dual-chamber, and single-chamber atrial pacemakers are equally effective with no difference in long-term mortality, HF, or stroke (190). However, the incidence of AF and risk of pacemaker reoperation is higher with single-chamber atrial pacing (190). While implantation of an additional lead with dual-chamber pacemakers increases procedural risk, this must be weighed against the risk of future development of AV block, which would require reoperation to place an additional ventricular lead (2, 190). The risk of developing AV block after pacemaker implantation can be difficult to predict, with reported rates ranging anywhere from 3 to 35% within 5 yr of follow-up (2). Thus dual-chamber pacemakers are generally used in SND with normal AV conduction unless there is reason to avoid a right ventricular (RV) lead or if strict maintenance of AV synchrony is not deemed necessary (e.g., frailty, limited functional capacity, poor short-term prognosis) (2). Currently, dual-chamber devices are by far the most used pacemaker for SND in the United States, accounting for more than 80% of device implants (191). When dual-chamber pacemakers are used, efforts to minimize RV pacing should be made to further reduce the risk of AF and the long-term deleterious effects of continuous ventricular pacing (192).

There are certain risks associated with permanent pacemaker implantation, including complications associated with the implantation surgery itself as well as complications with the pacemaker over time (193). Pacemaker-related complications occur in up to one in five patients at 5 years and are mainly attributed to the leads (lead dislodgement, fracture, malfunction) or generator pocket (infection, hematoma) (194). Leadless pacemakers are a new technology designed to reduce lead and pocket-related complications by combining the pulse generator and electrode in a self-contained unit that is directly implanted into the RV (195). While only capable of ventricular pacing, leadless pacemakers are a reasonable alternative to atrial or dual-chamber pacing for SND when implantation of a transvenous system is considered difficult or high risk (poor venous access, need to preserve veins for hemodialysis, high risk of infection) (196). Furthermore, leadless pacemakers may be preferable in patients with symptomatic SND that is short-lived or infrequent and the pacing burden is anticipated to be low (2).

Temporary Pacemakers

Temporary pacing can be used acutely to manage bradycardia causing hemodynamic instability refractory to pharmacological therapy. These situations can include prolonged sinus pauses, life-threatening ventricular arrhythmias precipitated by bradycardia, or other severe bradycardia attributable to a reversible cause with the goal to avoid permanent pacemaker implantation (2). However, the use of temporary pacing in SND is uncommon as SND is generally an indolent chronic process that is rarely acutely life-threatening (2). The two main types of temporary pacemaker therapy are transcutaneous cardiac pacing (TCP) and transvenous temporary cardiac pacing (TV-TP).

Transcutaneous cardiac pacing.

TCP is a temporary pacemaking method involving the delivery of electric pulses to the patient’s chest to cause contractions of the heart. It is commonly used during life-threatening cardiac emergencies, such as severe symptomatic or hemodynamically unstable bradyarrhythmias, ventricular, and supraventricular tachycardia, or when episodes of such severe arrhythmias are predicted to occur (197). In patients with bradycardia who are not in full cardiac arrest, the use of TCP has been shown to be an effective measure for increasing HR and is generally tolerable by most patients (198). Studies have also shown mild improvement in survival with prehospital TCP in nonasystolic patients with symptomatic bradycardia (199) although no such benefit is seen in patients with full cardiac arrest (200). TCP is most commonly used as a temporizing measure for pacing until a temporary transvenous pacing wire or permanent pacemaker can be implanted.

Transvenous temporary cardiac pacing.

Transvenous temporary cardiac pacing (TV-TP) is typically performed by inserting a pacing wire into the RV from a central venous access site. The RV is generally targeted because of its ease of access and efficacy for SND or AV block, although temporary pacing of the RA can be performed when strict AV synchrony is necessary and there is no AV block (2). Early studies of TV-TP showed high rates of malfunction of up to 43%, regardless of insertion site (201). Venous thrombosis has also been known to occur in 30 to 45% of patients as a result of electrode insertion (202). Metkus et al. (203) more recently reported that among a large cohort of patients who received TV-TP, 37.9% eventually underwent permanent pacemaker implants. The risk of infection following permanent pacemaker implant is increased in patients with TV-VP before pacemaker implant. Given the risks associated with TV-VP, its use is reserved for severe SND with hemodynamic compromise as a bridge toward permanent pacemaker implant or until the bradycardia resolves (2).

Pharmacological Treatments

A pharmacological approach can be used for the acute management of severe bradycardia attributable to SND. According to current guidelines, atropine, isoproterenol, aminophylline, or theophylline can be used to acutely treat bradycardia that may result from SND (2). Atropine works to block the muscarinic acetylcholine receptor and subsequently reduces the activity of GIRK channels or I_K,Ach associated with SND, thereby increasing SAN automaticity (2). Atropine is an effective temporary treatment for bradycardia with minimal risk of ischemia or arrhythmia and is the first-line pharmacological agent in most cases (2). Atropine should be avoided in cases of infranodal conduction block because of the potential worsening of block and bradycardia. Isoproterenol is a nonselective β-receptor agonist with chronotropic and inotropic effects. Primarily used during electrophysiology studies, isoproterenol is useful for increasing HR in unstable bradycardias without vasoconstrictor effects (2). However, isoproterenol should be avoided in cases of suspected coronary ischemia as its nonselective beta-agonist effects increase myocardial oxygen demand while decreasing coronary perfusion (2). In addition, in some patients, isoproterenol may have a vasodilatory effect or a weak effect on HR response (2). Aminophylline and theophylline competitively antagonize adenosine receptors, increasing cardiac automaticity and exerting positive chronotropic effects (204). These drugs have been used for both the acute and chronic management of SND (10). In particular, these drugs are useful for the management of bradycardia following heart transplantation or spinal cord injury, which are refractory to atropine due to cardiac autonomic denervation (2). In SND cases where direct association of symptoms to bradycardia is unclear, a trial of oral theophylline can be considered to help correlate symptoms (2). In this setting, oral theophylline is seen to increase resting HR, decrease the frequency of sinus pauses, and improve subjective symptoms in a majority of patients, although to a lesser degree compared with permanent pacing (205, 206). In patients with symptomatic SND who decline permanent pacemaker implantation or are not candidates otherwise, oral theophylline can also be considered as an alternative treatment (2). While generally well tolerated, caution should be given when using theophylline in patients with tachycardia-bradycardia because of the increased frequency of atrial tachyarrhythmia and in patients with structural heart disease because of the increased risk of ventricular arrhythmias (10).

RESEARCH OPPORTUNITIES

With the goal of developing better and alternative therapeutic strategies for SND, future research is needed to better understand the causative cellular and molecular mechanisms of SND. In this section, we provide our perspective and highlight a few areas of interest for future investigations.

Aging-Related Processes in SND Pathogenesis

It is known that SAN function and pacemaking activity decline with age. There is evidence implicating age-dependent decreases in the expression of genes involved in the membrane clock and Ca²⁺ clock (45, 49). However, whether the function of certain ion channels and Ca²⁺ handling proteins presents age-dependent impairment requires experimental investigation. Moreover, whether the relative contribution of the membrane clock and Ca²⁺ clock to pacemaking activity is altered with aging deserves to be addressed. Naturally, aging is associated with increased atrial fibrosis, more inflammation, mitochondrial dysfunction, epigenetic alterations, and genomic instability. How these processes alter the function of pacemaker cells and the structure of the SAN is another exciting area to be explored.

Sex Impact on SND

There is some evidence to suggest that familial SND may represent a strong male predominance (207). As several risk factors of SND such as AF and HF present sex-dependent differences (208, 209), future studies are needed to understand the impact of sex in the pathogenesis of SND.

Genetic and Epigenetic Mechanisms Underlying SND

Although the genetic association between SHOX2, LMNA, and SND exists, the underlying mechanisms are currently lacking. Examination of the Shox protein level in SAN tissue in humans with and without SND is a prerequisite for future projects in examining the role of Shox2 in SND development. Lamins are very important in many different biological processes. To understand the specific contribution of lamins in SAN and CCS homeostasis and dissect the molecular mechanisms underlying LMNA-associated SND or cardiac conduction disorders, SAN pacemaker cell-specific or CCS-specific LMNA mutant alleles are needed. Findings in these areas can also shed a light on the epigenetic regulation’s involvement in SND development.

Cellular Model of Congenital SND

The association between genetic variants and SND is present. To assess the pathogenicity of any given variants, a cellular model system that can faithfully represent SAN pacemaker cells would be necessary. One classic approach is to reprogram cardiomyocytes cell lines (e.g., HL-1 cells) into SAN-like cells in a dish (210–212) and then introduce the mutated genes by transfection in the heterologous system, followed by electrophysiological assessment. As the techniques in differentiating human iPSCs become mature and commonly available (213), another opportunity is to develop human pacemaker-like cells with the human iPSCs derived from the SND-variant carriers. The advantages of iPSCs in cardiac electrophysiology research include 1) they can be easily scaled up, 2) they can be used for drug screening, and 3) they are helpful in gaining initial insights into monogenic disease (213, 214). While protocols for iPSC differentiation into cardiomyocytes have been optimized over the last decade and become widely available, protocols to differentiate iPSCs into pacemaker cells are still needed in the field. Based on developmental biology, transcription factors such as Tbx3, Tbx18, and Shox2, which are essential for SAN development, are ideal candidates for inducing iPSC-pacemaker cell differentiation. A transgene-independent method of generating SAN-like pacemaker cells from human iPSCs was described recently (215, 216). Through stage-specific modulation of signaling pathways including activation of bone morphogenic protein, retinoic acid, and inhibition of fibroblast growth factor, Protze et al. (215, 216) have shown success in producing enriched populations of SAN-like pacemaker cells with high efficiency. These SAN-like pacemaker cells exhibit typical pacemaker electrophysiological characteristics and are capable of pacing host tissue.

Biological Pacemaker Development

The electronic pacemaker device is the current mainstay treatment for SND. As discussed earlier, limited battery life, infection, and other device-related complications are the major disadvantages associated with the implantable pacemakers. There are ongoing efforts to reproduce a biological pacemaker as an alternative to the electronic pacemaker device (217, 218). Many different gene-based and cell-based approaches to produce biological pacemakers have been described over the years. However, it remains a challenge to translate these studies into clinical trials as the dose of gene expression and the location of delivery need to be optimized and perhaps individualized for SND with different etiologies. For a detailed review on biological pacemaker development, we refer you to a recent review article by Cingolani et al. (219).

SAN Mechanosensitivity

Our understanding of the molecular mechanisms of SAN mechanosensitivity remains limited, with further studies needed to identify the precise pathways implicated in this process. Identified components may offer a novel therapeutic target for the treatment of SND by halting maladaptive stretch-induced changes, thereby normalizing SAN function (18). Changes in mechanical loading observed with HF and AF may offer a mechanistic link between the development of SND and these comorbidities. Finally, consideration of SAN mechano-sensitivity is critical for future mechanistic studies of SAN function. Since many in vitro studies analyze SAN function in the absence of a mechanical load, these nonphysiological conditions may confound extrapolation of findings to in vivo SAN function and further methods to accurately model mechanical stress are needed for future SND studies.

Potential Role of Nonnodal Cells

Recent work has revealed the presence of numerous nonnodal cells within the SAN, including S100 calcium-binding protein B (S100B)⁺/glial fibrillary acidic protein (GFAP)⁺ peripheral glial cells, S100B⁺/GFAP^- interstitial cells, telocytes, as well as neuronal fibers projecting from the epicardial ganglionated plexuses directly into the SAN itself (220). S100B⁺ interstitial cells within the SAN have been shown to modulate rhythmic Ca²⁺ activity of pacemaker cells and lead to HR variability (220). The nonnodal cells form a unique meshwork of interconnected cells, which are hypothesized to help integrate the multitude of heterogeneous behaviors across the pacemaker cell network to produce the net impulse activity that exits the SAN, in a manner reminiscent of brain neuronal networks (220, 221). This “neuromimetic” framework within the SAN offers a novel conceptualization of SAN structure and highlights the potential role of nonnodal cells in maintaining normal SAN function. Identification of these previously undefined mechanisms of arrhythmia may offer yet another therapeutic target for the treatment of SND.

CONCLUSIONS

Significant advances have been made over the past two decades in understanding the pathophysiology of SND. Identification of genetic abnormalities associated with familial SND has helped the field to gain the current knowledge on the cellular and molecular mechanisms of SND. Nevertheless, abundant research opportunities regarding the pathogenesis of SND remain. Mechanistic insights into how age-dependent changes in metabolism and epigenetic regulation promote the pathogenesis of nonfamilial SND are needed. Biological pacemaker development remains challenging but is an area with tremendous potential as the field of gene therapy and regenerative medicine for cardiovascular diseases is moving forward.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01HL136389, R01HL163277, and R01HL147108 (to N. Li); American Heart Association Established Investigator Award 936111 (to N. Li); and Patient-Centered Outcomes Research Institute research support (to M. G. Chelu).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.K., M.G.C., and N.L. conceived and designed research; P.M., J.A.K., S.K., T.L., and M.S. analyzed data; J.A.K., S.K., T.L,. and N.L. interpreted results of experiments; P.M., J.A.K., and N.L. prepared figures; P.M., J.A.K., S.K., and M.S. drafted manuscript; P.M., J.A.K., S.K., T.L., M.G.C., and N.L. edited and revised manuscript; P.M., J.A.K., S.K., T.L., M.S., M.G.C., and N.L. approved final version of manuscript.