Basic Science for Clinicians: The Cardiac Conduction System

Introduction

The human heart beats 2.5 billion times during a normal lifespan, a feat accomplished by cells of the cardiac conduction system (CCS). The functional components of the CCS can be broadly divided into the impulse generating nodes and the impulse propagating His-Purkinje system. Human diseases of the conduction system have been identified that alter impulse generation, propagation, or both. CCS dysfunction is primarily due to acquired conditions such as myocardial ischemia/infarct, age-related degeneration, procedural complications, and drug toxicity. Inherited forms of CCS disease are rare, but each new mutation provides invaluable insight into the molecular mechanisms governing CCS development and function. Applying a multidisciplinary approach, which includes human genetic screening, biophysical analysis, and transgenic mouse technology, has yielded a broad array of gene families involved in maintaining normal CCS physiology (figure 1). In this review, we discuss gene families that have been implicated in human CCS diseases of rhythm, conduction block, accessory conduction, and development (Table I). We also investigate evolving therapeutic strategies that may serve as adjuvant or replacement therapy to current implantable pacemakers.

Figure 1. Cardiac conduction system cell.

Genes identified in human cardiac conduction system disease are highlighted.

Table I.

Genetic Basis of Conduction System Disease

Gene Name	Protein	Associated Condition	Conduction Defect	Mechanism	References
Ion Channels
SCN5A	Nav1.5	Brugada syndrome, LQT3	SSS, PCCD, atrial standstill, AV block, bundle branch block	Reduced cardiac excitability and slowed conduction. Loss of depolarizing current in peripheral SAN cells	1–4, 5–8
SCN1B	Scn1b	Brugada syndrome	Progressive cardiac conduction defect		9
KCNJ2	Kir2.1	Andersen-Tawil syndrome	AV block, bundle branch block	Prolongation of action potential duration and reduced cardiac excitability	10, 11
HCN4	HCN4		Sinus bradycardia	Reduction in pacemaker current (If)	12–14
Ca2+ Handling Proteins
RYR2	Ryr2	CPVT	SND, AV block, atrial standstill	Altered Ca2+ handling likely affecting the Ca2+ clock	15, 16
CASQ2	Calsequestrin	CPVT	Sinus bradycardia		17
Gap Junction Proteins
GJA5	Cx40		Atrial standstill	Impaired myocyte coupling resulting in slowed conduction	3
Transcription Factors
TBX5	Tbx5	Holt-Oram Syndrome	Sinus bradycardia, AV block, bundle branch block	Specification defect of the AVN and the ventricular CCS	18
NKX2–5	Nkx2–5		AV block, bundle branch block		19
Nuclear Membrane Proteins
LMNA	Lamin A/C	Emery-Dreifuss muscular dystrophy	AV block	Altered nuclear stress mechanics and hyperactivation of MAPK signaling	20
Membrane Adapter proteins
ANKB	Ankyrin-B		Sinus bradycardia and heart rate variability	Altered ion channel and transporter expression and trafficking	21
Metabolic Regulators
PRKAG2	g- subunit of AMPK	Glycogen storage disease	WPW, AV block	Altered ventricular myocyte energetics leading to glycogen engorged vacuoles and disruption of annulus fibrosus	22, 23
TGF-b Superfamily
BMP2	Bone morpho-genetic protein 2		WPW	Maturation defects of annulus fibrosus	24
Spliceopathies
DMPK 3’UTR		Myotonic dystrophy type I (Steinert’s disease)	AV block, intraventricular conduction disease	Toxic mutant RNA and altered splicing factor function	25

Diseases of Automaticity

The human sinoatrial node is a cresent-shaped, intramural structure with its head located sub-epicardially at the junction of the right atrium and the superior vena cava and its tail extending 10–20mm along the crista terminalis.²⁶ The SAN has complex 3-D tissue architecture with central and peripheral components made up of distinct ion channel and gap junction expression profiles.²⁷ Central and peripheral cells have different action potential characteristics and conduction properties (figure 2).²⁷ Experimental and computational models have demonstrated that SAN heterogeneity is necessary to maintain normal automaticity and impulse conduction.^28–30

Figure 2. Electrophysiologic heterogeneity of the sinoatrial node.

The central SAN, the site of dominant pacemaking, is electrotonically insulated from the hyperpolarizing atrial myocardium through the differential expression of connexins and ion channels. Peripheral SAN cells are electrophysiologically intermediate between central cells and atrial cardiomyocytes.

Pacemaker automaticity is due to spontaneous diastolic depolarization of phase 4, which depolarizes the membrane to threshold potential generating rhythmic action potentials. The current paradigm of SA node automaticity has been modeled as two clocks that function in concert, the “membrane voltage clock” and the “calcium clock”. The membrane voltage clock is produced by the net dysequilibrium between the decay of outward potassium currents (I_K) and the activation of inward currents that include but is not limited to, background sodium-sensitive current (I_{b Na}), L- and T-type calcium currents (I_Ca,L, I_Ca,T), sustained inward (I_st) current, and hyperpolarization-activated current (I_f) (figure 2).^{27, 31–33}

The sub-sarcolemmal calcium clock contributes to SAN diastolic depolarization through the spontaneous, rhythmic release of Ca²⁺ from the sarcoplasmic reticulum (SR) via the ryanodine type 2 receptor (RYR2).³⁴ The local intracellular calcium (Ca_i) elevations drive the sodium-calcium exchange current (I_NCX) to substitute one intracellular Ca²⁺ for 3 extracellular Na⁺. The net gain in positive charge results in membrane depolarization.³⁵ The elevation of intracellular Ca²⁺ occurs in the latter third of diastolic depolarization and is sensitive to β-adrenergic stimulation.³⁶ Human mutations affecting the voltage clock (SCN5A and HCN4), calcium clock (RYR2 and CASQ2), or both mechanisms (ANKB) have been identified that negatively impact sinus node function.^{37, 38}

SCN5A

SCN5A encodes the cardiac sodium channel (Nav1.5), which generates the fast sodium current, I_Na. Nav1.5 dictates the amplitude and slope of phase 0 of the cardiac action potential and therefore impacts conduction velocity. The expression of SCN5A is restricted to the peripheral SAN and absent in the central SAN, the site of dominant pacemaking (figure 2).

Although the role of I_Na in SAN automaticity is not completely understood, several SCN5A mutations have been identified in familial sick sinus syndrome (SSS). SSS can manifest as sinus bradycardia, sinus arrest, sinus exit block, and atrial standstill. SCN5A mutations indirectly affect central SAN automaticity through their actions on peripheral nodal cells. More than 200 mutations have now been identified in SCN5A that cause a collection of cardiac diseases that include long QT3, Brugada syndrome, progressive cardiac conduction defect (PCCD), and congenital SSS. To date, 14 SCN5A mutations have been linked to inherited forms of SSS.^{27, 39}

Benson et. al. identified 6 SCN5A mutations in 3 kindreds with inherited SSS. The probands exhibited compound heterozygosity of the 6 SCN5A alleles (G1408R+P1298L, T220I+R1623X, delF1617+R1632H). Biophysical evaluation revealed that 2 of the 6 mutations resulted in non-functional channels while T220I, P1298L, and delF1617 demonstrated reduced I_Na current density and a hyperpolarizing shift in voltage-dependent inactivation.¹ Smits et al. identified a novel SCN5A mutation, E161K, in two kindreds with symptomatic sinus node dysfunction, generalized conduction disease, and Brugada syndrome. Whole-cell patch clamp revealed a three-fold reduction in I_Na current density and a positive shift in voltage-dependent activation.⁴⁰ Both a negative shift in voltage-dependent inactivation and a positive shift in voltage-dependent activation result in narrowing of the I_Na current window.

A proposed mechanism of how peripherally expressed SCN5A mutations affect central SAN automaticity was computer simulated by Butters et al. SCN5A mutants (T220I, P1298L, delF1617, and E161K) were incorporated into single cell models and into a 2-D model of intact, rabbit SAN-atrium.⁴¹ Single cell simulations showed that SCN5A mutants reduced the pacemaking rate of peripheral cells with little impact on central nodal cells. In contrast, intact tissue simulations with the SCN5A mutants slowed the sinus rate within the central SAN. The atrium imposes a significant hyperpolarizing load on the SAN, which is counterbalanced by SCN5A expression in peripheral cells.^41–43 Therefore, reduced I_Na in the peripheral SAN exposes the central SAN to more hyperpolarized potentials, slowing pacemaking rate. In addition, impaired action potential conduction was evident across the SAN-atrium predisposing to sinus exit block, a common feature of familial SSS.⁴¹

HCN4

The pacemaker or “funny” current, I_f, is generated by the hyperpolarization-activated cyclic nucleotide-gated (HCN) channel. The term “funny” current stems from the unique characteristics of HCN channels, which include permeability to K⁺ and Na⁺, activation at hyperpolarized membrane potentials, and modulation by cyclic AMP (cAMP). The biophysical properties of HCN channels make them ideally suited to function as modulators of the pacemaker potential. First, because they are activated at hyperpolarized membrane potentials (between −65 to −40mV), their slow inward current contributes to diastolic depolarization, and second, the main cardiac HCN channel (HCN4) is cAMP responsive allowing I_f to be modulated by autonomic stimulation.^{44, 45}

Mutations in HCN4, the predominant isoform in the SAN, have been identified in patients with SSS.^{32, 45} Using candidate gene strategies, Schulze-Bahr et al. identified a single nucleotide deletion in HCN4 (1631delC) in a patient with sinus bradycardia and chronotropic incompetence. The 1631delC mutant lacks the cyclic-nucleotide binding domain making it unresponsive to cAMP.¹² Additional HCN4 mutations were identified in two families with asymptomatic sinus bradycardia. The two missense mutations, S672R and G480R, resulted in channels that activate at more hyperpolarized voltages generating smaller currents during diastolic depolarization.¹³

Investigations into animal models of HCN channel deficiency have shed light on their role in I_f current generation. HCN4 knockout mice were embryonic lethal at E10.5 to E11.5 and demonstrated a cardiomyocyte maturation defect. HCN4^−/− embryonic cardiomyocytes exhibited an 85% reduction in I_f and were chronotropically unresponsive to cAMP.^{46, 47} To study the loss of HCN4 expression in adult mice, tamoxifen-inducible mouse models were generated.^{48, 49} The loss of HCN4 in SAN cells resulted in a 75% reduction in I_f current density. Instead of manifesting sinus bradycardia, these mice developed sinus pauses up to 300–500ms. Chronotropy was also preserved. HCN4^−/− SAN cells exhibited normal automaticity, however, diastolic potentials tended to drift to hyperpolarized levels where spontaneous pacemaker activity would cease. Taken together, these results suggest that HCN4 is indispensible for SAN development and chronotropy during embryogenesis, but may play only a backup role to counter membrane hyperpolarization in the adult SAN.^47–49 These results are consistent with the benign bradycardic phenotype seen in the two families with HCN4 mutations reported by Milanesi and Nof.^{13, 14}

The preservation of chronotropic competence in human and experimental models of HCN4 deficiency is suggestive of alternative mechanisms of SAN automaticity and sympathetic responsiveness. The first implication of Ca_i in diastolic depolarization was reported by Rubenstein et al. when ryanodine treatment was found to slow automaticity in subsidiary pacemakers in cat right atria.⁵⁰ A mechanistic link between Ca_i and SAN diastolic depolarization was subsequently reported by Huser et al. who proposed that subsarcolemmal Ca²⁺ transients released from the SR (i.e. Ca²⁺ sparks) lead to activation of the Na-Ca exchanger (NCX) resulting in membrane depolarization.³⁴ Bogdanov and Lakatta confirmed this mechanism in isolated rabbit SAN cells and directly linked Ca²⁺ sparks to late diastolic depolarization via the NCX.³⁵ Vinogradova et al. then showed that β-adrenergic augmentation of rabbit SAN chronotropy is dependent on SR Ca²⁺ release as treatment with ryanodine blunted the pacemaker rate response to isoproterenol.^{36, 51} Recent work by Joung et al. examined the effect of adrenergic stimulation on intact canine SAN preparations using dual optical mapping of transmembrane potential (V_m) and intracellular calcium (Ca_i). Adrenergic stimulation resulted in a robust increase and superior shift in late diastolic Ca_i elevations (LDCAE) that colocalized with the primary pacemaking site.³⁸

RYR2 and CASQ2

It is therefore interesting that mutations in SR calcium handling proteins are associated with sinus node dysfunction. Sinus bradycardia is a common feature of catecholaminergic polymorphic ventricular tachycardia (CPVT), an inherited arrhythmia syndrome characterized by bi-directional or polymorphic VT induced by adrenergic stress in the absence of structural heart disease. Two CPVT-linked genes have been identified in humans, the cardiac ryanodine receptor, RYR2, and cardiac calsequestrin, CASQ2. The cardiac ryanodine receptor is the main calcium release channel on the sarcoplasmic reticulum (SR) in cardiomyocytes. Calsequestrin is the major calcium storage protein in the SR and complexes with RYR2. Mutations in these proteins result in altered calcium release characteristics from the SR leading to catecholamine-induced spontaneous SR Ca²⁺ release resulting in delayed after-depolarizations and triggered activity.^{52, 53}

Postma et al identified 13 RYR2 missense mutations in 12 families with CPVT. A total of 54 family members were found to be carriers and all had resting sinus bradycardia off medications.¹⁶ A variant of CPVT with extended features of dilated cardiomyopathy, SND, progressive AV block, atrial fibrillation, and atrial standstill was recently reported. Linkage analysis and long-range PCR revealed an in-frame, genomic deletion involving RYR2 exon3.¹⁵ Sinus bradycardia is also a characteristic feature of CPVT due to CASQ2 mutations. Postma et al. identified three nonsense mutations in CASQ2 (a nonsense R33X, a splicing 532+1 G>A, and a 1-bp deletion, 62delA) in 3 kindreds with CPVT; all manifested sinus bradycardia on baseline ECG.¹⁷ The association of sinus bradycardia with mutations in SR calcium handling proteins lends further support for the calcium clock hypothesis in sinoatrial pacemaking, however the mechanism of SND in CPVT will need further study.

ANKB

Mutations in Ankyrin-B (ANKB) are associated with long QT type 4⁵⁴ and familial sinus node dysfunction²¹. ANKB is required for the proper targeting of ion channels and transporters to specific membrane domains. Mouse models haploinsufficient for AnkB develop profound sinus bradycardia and resting heart rate variability. Biochemical analysis revealed reduced expression and improper targeting of the Na/Ca exchanger (NCX1), Na/K ATPase (NKA), IP3 receptor (IP3R), and Ca_v1.3. I_Ca,L and I_NCX were significantly reduced in AnkB+/− SAN cells.²¹ Therefore, ANKB mutations likely cause sinus node dysfunction by reducing ionic currents involved in both clock mechanisms.

Diseases of Conduction Block

Conduction block can occur at any level of the CCS and can manifest as SA exit block, AV block, infra-Hisian block, or bundle-branch block. Impaired conduction can be caused by ion channel defects that alter action potential shape or due to defective coupling between cardiomyocytes. Inherited defects in cardiac conduction have been linked to mutations in SCN5A and SCN1B (both affect phase 0) and KCNJ2 (affects phase 3 and 4).

The cardiac sodium channel consists of the pore-forming α-subunit (encoded by SCN5A) and a modulatory β-subunit (encoded by SCN1B). The α-subunit contains a voltage sensor that allows for rapid activation in response to membrane depolarization. After depolarization, the sodium channel undergoes a period of inactivation, where it is refractory to further impulses. SCN5A requires membrane repolarization in order to relieve the inactivated state. The inward rectifier potassium channel, Kir2.1, encoded by KCNJ2, maintains the resting membrane potential. Therefore, proper functioning of Nav1.5 and Kir2.1 is necessary for normal cardiac excitability.

SCN5A

Progressive cardiac conduction defect (PCCD), or Lev-Lenegre disease, is characterized by age-related, fibrosclerotic degeneration of the His-Purkinje system.⁶ Impulse propagation through the proximal ventricular conduction system progressively declines resulting in bundle branch blocks and eventually, complete AV block. An inherited form of Lev-Lenegre disease is associated with loss of function mutations in SCN5A and can exist alone or as overlap syndromes with Brugada or LQT3.⁶ Inherited PCCD is associated with a high risk of complete AV block and Stoke-Adams syncope without ventricular dysrhythmia.⁷ Schott et al identified a mutation in SCN5A that cosegregates with Lenegre disease in a large French family. ⁸ Affected individuals had variable degrees of conduction block requiring pacemaker implantation in four family members due to syncope or complete heart block. Linkage analysis and candidate gene sequencing identified a T->C substitution at position +2 of the donor splice site of intron 22 (IVS22+2 T->C), which results in a mutant lacking the voltage-sensitive segment.⁸ Functional analysis demonstrated no transient inward sodium current in response to depolarization, consistent with a loss-of-function mutation.⁶

SCN1B

The majority of Brugada and conduction disease patients do not have SCN5A mutations. Therefore, modifiers of Nav1.5 expression or function have become the target of candidate gene sequencing approaches. Watanabe et al. identified SCN1B mutations in 3 families with conduction disease with or without Brugada syndrome. Co-expression of mutant β-subunits with Nav1.5 resulted in diminished sodium current.⁹

KCNJ2

Mutations in KCNJ2 have been found in a rare, autosomal dominant condition called Andersen-Tawil syndrome (ATS), characterized by periodic paralysis, dysmorphic features, polymorphic VT, and cardiac conduction disease.^{10, 11} ECG evaluation on 96 ATS patients from 33 unrelated kindreds revealed conduction defects at multiple levels from the AVN to the distal conduction system.⁵⁵ Cardiomyocytes expressing a dominant-negative subunit of Kir2.1, exhibited a 95% reduction in I_K1 resulting in significant action potential prolongation. Mouse models of ATS exhibited a slower heart rate and significant slowing of conduction.^{56, 57}

Diseases of Accessory Conduction

Wolff-Parkinson-White (WPW) syndrome is characterized by preexcitation of ventricular myocardium via an accessory pathway (bundle of Kent) that bypasses the normal slow conduction through the AVN. Ventricular pre-excitation is common with a disease prevalence of 1.5–3 per 1000 people.^{22, 58} Histological evaluation of Kent bundles resected from human subjects displayed features of typical ventricular myocytes with expression of Cx43.⁵⁹ The expression of high conductance gap junctions in bypass tracts enables them to pre-excite ventricular myocardium, manifesting as a short PR and a slurred QRS complex, or “delta wave”, on the ECG. The vast majority of WPW cases are sporadic and the underlying mechanism remains unknown; however, rare inherited forms have been reported. Vidaillet et al. determined that 3.4% of probands with WPW had one or more first-degree relatives with accessory conduction.⁶⁰

PRKAG2

A familial form of WPW with an autosomal dominant mode of transmission was identified in two families. Thirty-one affected individuals had evidence of pre-excitation and cardiac hypertrophy. A missense mutation in PRKAG2 was identified that results in a constitutively active form of the γ2 regulatory subunit of AMP-activated protein kinase (AMPK).^{22, 23} Under normal conditions, AMPK responds to energy-depleted states by increasing glucose uptake and promoting glycolysis. Transgenic mice expressing a heart-restricted, constitutively active mutant, PRKAG2^N488I, recapitulated the human WPW phenotype of cardiac hypertrophy, pre-excitation, and conduction defects. The predominant histological finding was ventricular myocyte engorgement with glycogen-laden vacuoles. The disruption of the annulus fibrosus by vacuolated ventricular myocytes resulted in the pre-excitation phenotype.⁶¹ Using a mouse model of reversible glycogen-storage defect, Wolf et al. demonstrated that the cardiomyopathy and CCS degeneration seen in PRKAG2^N488I mice were reversible processes.⁶²

BMP2

Lalani et al. reported a novel WPW syndrome associated with microdeletion of the bone morphogenetic protein-2 (BMP2) region within 20p12.3 that is characterized by variable cognitive deficits and dysmorphic features.²⁴ The BMPs are members of the transforming growth factor-β (TGFβ) superfamily and play a critical role in cardiac development. Mice with cardiac deletion of BMP receptor, type Ia (Bmpr1a) were embryonic lethal before E18.5 due to abnormal development of trabecular and compact myocardium, interventricular septum, and endocardial cushion.⁶³ More restricted deletion of Bmpr1a in the AV canal resulted in defective AV valve formation and maturation defects in the annulus fibrosus resulting in pre-excitation.^{64, 65}

Diseases of CCS Development

Congenital heart disease is the most common form of birth defect affecting 1–2% of live births.⁶⁶ Arrhythmias may result from defective CCS specification/patterning, malformation or displacement of the conduction system, altered hemodynamics, prolonged hypoxic states, or postoperative sequelae.^{67, 68} Developmentally, the conduction system derives from myocardial precursor cells within the fetal heart.^69–71 The process by which conduction cells are specified or recruited into a “conduction” versus “working myocyte” lineage is determined by the regional expression of transcription factors.^69–74 The main transcription factors identified in human CCS development are the T-box and homeobox factors.

TBX5

Holt-Oram syndrome (HOS) is an autosomal dominant condition characterized by preaxial radial ray limb deformities (defects of the radius, carpal bones, and/or thumbs) and cardiac septation defects. The septal defects are typically ostium secundum atrial septal defects (ASD), muscular ventricular septal defects (VSD), and atrioventricular canal defects (AVCD). HOS patients manifest variable degrees of CCS dysfunction, such as sinus bradycardia and AV block, even in the absence of overt structural heart disease. In 1997, Basson et al. screened two families with HOS and identified mutations in the T-box transcription factor, TBX5. The T-box transcription factors can function as transcriptional activators or repressors and are known to be critical regulators of cardiac specification and differentiation.¹⁸ Seven TBX family members are expressed in the developing heart, three of which (TBX1, TBX5, TBX20) have been linked to human congenital heart disease.⁷⁵

Mice deficient in Tbx5 were embryonic lethal at E10.5 due to arrested development of the atria and left ventricle. Tbx5^+/− mice recapitulated the upper limb and cardiac manifestations of human HOS including the conduction abnormalities.^{72, 76} Significant maturation defects in the AVC and ventricular conduction system were present.⁷⁶ Moskowitz et al. demonstrated that Tbx5^+/− mice have maturation failure of the AVC manifesting as persistent AV rings around the tricuspid and mitral valve. Patterning defects were noted in the His bundle and bundle branches including complete absence of right bundle branch formation in some cases.⁷⁶ Expression of CCS-enriched markers, such as atrial natriuretic factor (ANF) and Cx40, were found to be significantly downregulated, implicating Tbx5 as a transcriptional activator of these genes. Tbx5 and the homeobox transcription factor, Nkx2–5, were found to act synergistically in upregulating ANF and Cx40 expression.⁷⁶

NKX2–5

NKX2–5, a member of the homeodomain family, plays a central role in CCS induction and maintenance. Members of the homeodomain family all share a conserved 60 amino acid DNA binding motif known as the homebox. These transcription factors are essential in organogenesis, dictating tissue specification and differentiation. Loss of the Nkx2–5 homolog, tinman, in the fruit fly results in failure of cardiogenesis.⁷⁷ The role of NKX2–5 in human CCS development was established when mutations were identified in four families with non-syndromic, congenital heart disease and AV conduction block.¹⁹ Pedigree analysis was consistent with an autosomal dominant mode of transmission, and the most common structural abnormality was secundum atrial septal defects (ASD). Other associated structural heart defects included ventricular septal defects, tetrology of Fallot, pulmonary atresia, redundant mitral valve, left ventricular hypertrophy, and subvalvular aortic stenosis.¹⁹ Electrocardiogram analysis revealed AV conduction defects that were not strictly dependent on underlying structural heart disease. Fourteen individuals required pacemaker implantation. Genome-wide linkage analysis and candidate gene sequencing identified 3 mutations in NKX2–5.¹⁹ Currently, there are over 60 NKX2–5 single-nucleotide substitutions that have been identified in congenital heart patients emphasizing its critical role in cardiogenesis and conduction system specification and maintenance.

Mice deficient in Nkx2–5 die in utero at E9–10 due to arrested cardiac development in the linear heart tube stage. The hearts of these mice undergo partial looping morphogenesis, lack endocardial cushions and trabeculae, and have underdeveloped AV canals.⁷⁸ Jay et al. reported complete absence of AVN primordial cells in another model of Nkx2–5 deficiency (Nkx2–5^neo/neo) in a background of minK-LacZ, a reporter mouse that labels the SAN, AVN, and proximal ventricular conduction system.⁷⁹ Heterozygous Nkx2–5 mice exhibit an overall reduction in the size of the CCS from the AVN to the distal Purkinje network. Histological evaluation of the Nkx2–5^+/neo AVN revealed that compact nodal cells (N region) that are connexin 40−/connexin 45+ (Cx40⁻/Cx45⁺) are distinctly absent while the nodo-His (NH) region that is Cx40⁺/Cx45⁺ remains intact.⁷⁹ Immunostaining for Cx40 revealed that the density of Purkinje fibers was reduced.⁷⁹

Surface electrograms of Nkx2–5^+/neo mice showed PR and QRS prolongation. Intracardiac electrograms were notable for a diminutive His depolarization amplitude consistent with reduced size of the His bundle, and for prolonged 1:1 and 2:1 AV cycle lengths and AVN effective refractory periods suggestive of AVN dysfunction. His-ventricular (HV) intervals were unchanged likely due to sufficient expression of Cx40. Similarities in AVN dysfunction between humans and mice with Nkx2–5 haploinsufficiency point to the conserved role of Nkx2–5 in the induction and maintenance of the mammalian CCS.^79–81

Nkx2–5, Tbx5, and the inhibitor of differentiation 2 (Id2) are now known to function together as a conduction system transcriptome specifying the murine ventricular CCS.⁷⁴ The co-expression of Nkx2–5 and Tbx5 in the developing VCS results in regionally-restricted expression of Id2, ANF, and Cx40. Id2 is believed to function as an inhibitor of muscle differentiation allowing specification toward a conduction lineage.⁷⁴ Combined haploinsufficiency of Nkx2–5 and Tbx5 (Nkx2–5^+/−/Tbx5^+/−) results in developmental failure of the ventricular conduction system with complete absence of the His bundle and bundle branches. Loss of CCS markers, like Cx40, in mutant hearts results in marked widening of the QRS complex on surface ECG.⁷⁴ A human correlate of compound heterozygosity of Nkx2–5 and Tbx5 has not yet been identified.

Conduction Disease Associated with Neuromuscular Disorders

Neuromuscular disorders represent a diverse collection of diseases that commonly present with cardiac involvement. Mutations have been identified in genes involved in the cytoskeleton, nuclear envelope, and mitochondrial function. Cardiac involvement typically manifests as dilated or hypertrophic cardiomyopathy, AV conduction defects, and atrial and ventricular dysrhythmias.⁸²

EMD and LMNA

Mutations affecting the nuclear envelope have been associated with significant CCS dysfunction. The inner membrane of the nuclear envelope is a highly organized structure, composed of integral membrane proteins and nuclear cytoskeletal proteins that function together in higher-order chromatin structure and transcriptional regulation. The lamins (A, B, and C) are an integral part of intermediate filament network that imparts structural rigidity to the inner nuclear membrane. Emerin, a member of the nuclear lamina-associated protein family, putatively mediates anchoring of chromatin to the cytoskeleton. Mutations in Emerin (EMD) or the Lamin A/C (LMNA) result in X-linked, Emery-Dreifuss muscular dystrophy (X-EDMD) and autosomal dominant EDMD (AD-EDMD)²⁰, respectively. Individuals with EDMD develop progressive skeletal muscle weakness in the first decade of life and cardiac involvement (dilated cardiomyopathy and AV block) in the second decade.^{82, 83}

Arimura et al engineered a mouse model of AD-EDMD by knocking-in an Lmna missense mutation (H222P) previously identified from a family with typical AD-EDMD. The mouse model faithfully recapitulated the human disease with Lmna^H222P/H222P mice exhibiting locomotive defects, dilated cardiomyopathy and CCS dysfunction. Telemetric evaluation of the mutant mice revealed PR prolongation and QRS complex widening.⁸⁴ A similar CCS defect was seen in mice haploinsufficient in the Lmna gene. Lmna^+/− mice exhibited sinus bradycardia with variable degrees of AVB. Histological evaluation of these mice revealed nuclear deformation and apoptosis in AVN cells.⁸⁵ Another engineered mouse line expressing Lmna^N195K, known to cause autosomal dominant, dilated cardiomyopathy with conduction disease (DCM-CD1) in humans,⁸⁶ exhibited high-grade AVB and complete heart block. Biochemical evaluation revealed reduced expression and mislocalization of Cx40 and Cx43 in mutant atrial tissue.⁸⁷ Desmin staining also revealed structural defects of the sarcomere and intercalated discs.⁸⁷

Genome-wide expression profiling of Lmna H222P mouse hearts revealed significant increases in mitogen activated protein kinase (MAPK) signaling pathways.⁸⁸ Hyperactivation of MAPK pathways are associated with cardiomyopathy and CCS dysfunction. A significant increase of the activated forms of two MAPKs, JNK and ERK1/2, was noted in mutant hearts that predated the onset of overt or molecularly-defined cardiomyopathy.⁸⁸ Treatment of Lmna H222P mice with an inhibitor of ERK phosphorylation abrogated the increase in biomarkers of cardiomyopathy and restored ejection fraction to normal levels. These findings directly link MAPK hyperactivation to the cardiomyopathic phenotype in Lmna H222P mice.⁸⁹

Based on the phenotypes of these mouse models, lamin A/C appears to maintain the functional integrity of the CCS in two ways: i) protection of the nucleus against mechanical stress and ii) maintaining proper chromatin organization to ensure accurate gene expression, such as in connexin expression and MAPK signaling pathways.⁸³

DMPK

Myotonic dystrophy type I (DM1), or Steinert’s disease, is the most common muscular dystrophy in adults and is characterized by myotonia, insulin resistance, progressive skeletal muscle loss, cataracts, and cardiac conduction defects.²⁵ CCS dysfunction is seen in 70% of patients with DM1 and can manifest as sinus bradycardia, variable degrees of AV block, and intraventricular conduction delays.⁹⁰ The genetic basis of DM1 is an expanded (CTG)n repeat in the 3’ untranslated region (UTR) of the myotonic dystrophy protein kinase (DMPK) gene. Three molecular mechanisms linking the tri-nucleotide expansion with the cardiac conduction defects of DM1 have been proposed: i) haploinsufficiency of DMPK⁹¹, ii) under-expression of the neighboring homeodomain-encoding gene, Six5⁹², and iii) toxic accumulation of mutant RNA⁹³. Nuclear entrapment of toxic RNA leads to sequestration and altered activity of RNA-binding proteins, including proteins involved in RNA splicing like the muscleblind-like (MBL1) family.⁹³ Although defective splicing has been shown to be involved in myotonia and insulin resistance, the mechanism of conduction disease in DM1 remains unclear.

To explore the pathophysiologic role of the 3’UTR of DMPK in the DM1 phenotype, Mahadevan et al. created a transgenic mouse that expresses an inducible transcript encoding the DMPK 3’UTR with green fluorescent protein (GFP). They found that overexpression of the wild type DMPK 3’UTR was sufficient to induce the characteristic features of DM1, including myotonia, splicing defects, and CCS dysfunction. The authors postulated that transgenic overexpression of the wildtype DMPK 3’UTR with 5 (CUG)n repeats may be the pathogenetic molar equivalent of mutant 3’UTR with hundreds of (CUG)n repeats. Induction of the transgene first resulted in PR prolongation that progressed to complete heart block and sudden cardiac death. Interestingly, cessation of the inducing agent, doxycycline, resulted in reversion of the myotonic and some of the CCS defects. Animals exhibiting first-degree heart blocks were reversible, however, AV blocks beyond second degree were not reversible.⁹³

Future directions

Linkage analysis with positional cloning has been a highly effective means of identifying gene mutations within kindreds of monogenic disease. More than 1000 genes have been identified using this approach, including those in this review. With the sequencing of the human genome, the promise of identifying genetic causes of complex, multifactorial diseases is becoming more of a reality. One major step in this direction was the development of genome-wide association studies (GWAS).⁹⁴

The GWAS is a test of association between a disease and genetic markers that span the entire genome. The technique relies on the fact that variance at one locus predicts with high probability variance of an adjacent locus due to linkage dysequilibrium. In other words, there is non-random cosegregation of a series of genetic markers that are close together in the genome. This cluster of linked markers is known as a haplotype. The first study of haplotype structure within four populations (Yoruban, Northern/Western Europeans, Chinese, and Japanese) was published in Nature in 2005 by the International HapMap Consortium. Their work reported that individual genetic markers (single nucleotide polymorphisms, or SNP’s) predict adjacent markers typically with a resolution of about 30,000 bp. Considering the human genome is ~3×10⁹ bp, they projected that <500,000 SNP’s would be needed to survey the entire genome for all common genetic variability.^{94, 95}

Genome-wide association studies have now been used to identify genetic variants that influence ECG parameters in different populations. Intermediate parameters, such as heart rate or PR interval, were used as surrogate markers of disease for two reasons: i) they have an association with cardiovascular morbidity and atrial fibrillation and ii) they have tighter associations with gene variants than the actual disease. Holm et al reported several genome-wide associations using a cutoff P value <1.6×10⁻⁹. One locus harboring MYH6 was associated with HR, 4 loci (TBX5, SCN10A, CAV1, and ARHGAP24) were associated with PR interval, and four loci (TBX5, SCN10A, 6p21, and 10q21) with QRS duration. They went on to test these associations with individuals manifesting different arrhythmias in an Icelandic and Norwegian population. Correlations were found between atrial fibrillation and TBX5 and CAV1 (P=4.0×10–5 and P=0.00032, respectively), between advanced AVB and TBX5 (P = 0.0067), and between pacemaker implantation and SCN10A (P = 0.0029).⁹⁶

Similar loci were identified by two additional independent GWA studies in a European population and an Indian Asian population. Pfeufer et al reported 9 loci that were highly associated with PR interval (P < 5×10⁻⁸) from a meta-analysis of the CHARGE Consortium with over 28,000 European subjects. One locus had associations with 2 sodium channels (SCN10A and SCN5A) and 6 loci were near genes involved in cardiac development (CAV1-CAV2, NKX2–5, SOX5, WNT11, MEIS1, and TBX5-TBX3). Of these, SCN10A, SCN5A, CAV1-CAV2, NKX2–5, and SOX5 were found to be associated with atrial fibrillation.⁹⁷ Chambers et al also reported the association between SCN10A with PR interval in 6,543 Indian Asians. Physiologic testing of Scn10a deficient mice revealed shortened PR intervals in knockout mice with no significant difference in all other ECG and echocardiographic parameters.⁹⁸

The discovery of novel gene families associated with human conduction and arrhythmic diseases using genome-wide association studies is well underway. Identification of SCN10A by three independent groups studying different populations confirms the fidelity of this approach. Further experiments confirming the significance of these associations will need to be performed. In addition to identifying novel gene targets, this technique will also aid in the discovery of new associations with non-coding regions where new epigenetic modifiers and transcriptional/translational regulators, such as microRNA’s, will be identified.

Therapeutic Strategies

The current standard of care for symptomatic bradycardia due to conduction system disease is the implantation of an electronic pacemaker. Despite their success, electronic pacemakers have limitations, which include lead complications, finite battery life, potential for infection, lack of autonomic responsiveness, and size restriction in younger patients. These limitations have spurred on the development of biological pacemakers; the premise of which is to restore pacemaking activity using viral-based or stem cell-based gene delivery systems.⁹⁹ The identification and characterization of genes involved in generating pacemaker currents has allowed biological pacemaker technology to become a reality.

The restoration of sinus pacing rates can be achieved by modulating inward and outward currents to establish or increase the slope of diastolic depolarization in cardiac tissue. Increasing inward currents and/or decreasing outward currents increase the slope of diastolic depolarization, and therefore, pacing rate. Genes that have been investigated or are under current investigation include: i) β2-adrenergic receptor,^{100, 101} ii) dominant-negative Kir2.1 mutants,¹⁰² iii) adenylate cyclase type VI (ACVI),^{103, 104} and iv) HCN channels.¹⁰⁵ The β2-adrenergic receptor and adenylate cyclase VI both increase c-AMP levels leading to activation of endogenous HCN channels and calcium clock mechanisms. Although initial animal models using the β2-adrenergic receptor showed promise with transient increases in heart rate, the potential for proarrhythmia and the inability of this approach to establish de novo pacemaker activity limited its efficacy.¹⁰¹

Another approach focused on modifying ionic currents that convert working myocardial cells, which have relatively stable diastolic potentials, into cells with phase 4 diastolic depolarization. It was postulated that atrial and ventricular myocytes have the potential for automaticity, but that hyperpolarizing currents, such as I_K1, prevent diastolic depolarization by stabilizing the resting membrane potential. Miake et al. confirmed this hypothesis when they demonstrated that adenoviral delivery of a dominant-negative Kir2.1 construct into the LV of guinea pigs resulted in conversion of quiescent myocytes into pacemaker cells.¹⁰² Unfortunately, significant action potential prolongation limited the clinical utility of this treatment strategy.¹⁰²

Rosen and colleagues demonstrated that automaticity could be induced in quiescent myocardium using heterologous expression of HCN channels that produce the pacemaker current, I_f. Qu and Plotnikov et al. demonstrated that stable autonomous rhythms could be generated when adenovirus encoding HCN2 was injected into the left atrium¹⁰⁵ or left bundle branch¹⁰⁶ of a canine heart. In order to bypass the limitations of viral-based systems, such as host immune response, several groups reported the successful use of cell-based delivery systems. Plotnikov et al. reported the successful implantation of human mesenchymal stem cells (hMSC) expressing HCN2 in the LV of a canine model of AV block.¹⁰⁷ Dogs maintained stable ectopic pacemaker activity for over 6 weeks without the use of immunosuppression.¹⁰⁷ Human MSCs electronically couple to host myocardium through gap junctions; therefore, conditions with significant gap junction remodeling may affect the efficacy of this method.

Although standalone biological pacemakers may be far into the future, adjuvant biological pacemakers may find real-world utility for current deficiencies of electronic pacemakers, such as limited battery life and device infections. For example, biologicals used in conjunction with device therapy may be used to extend battery life, decreasing the frequency of generator changes. Transient injectable pacemakers may also function as bridge therapy after lead extraction of an infected device. The need for adjuvant biological pacemakers is clear, but continued refinement of gene and cell-based delivery systems will be necessary to make this technology a reality.⁹⁹

Conclusion

Though rare, inherited arrhythmias have become an invaluable tool in identifying the genetic determinants of CCS function. Each new mutation enhances our understanding and appreciation of the biochemical and structural complexity needed for cardiac impulse generation and propagation. This methodology is hampered however, by the relative scarcity of inherited conditions affecting the CCS. The addition of GWAS has broadened this search for novel genes beyond rare familial afflictions to include common, multi-factorial conditions. This exciting new frontier will hopefully bring to light the complex interplay of genes and genetic/epigenetic modifiers that influence the prevalence of common diseases. These genetic screens will ultimately yield a bevy of new gene targets for pharmaceutical or gene-based therapeutics of the future.