Calcium Revisited New Insights Into the Molecular Basis of Long-QT Syndrome

The first clinical description of long-QT syndrome (LQTS) occurred in 1957 when Drs Anton Jervell and Fred Lange-Nielsen postulated that syncopal/seizure episodes and high propensity for sudden cardiac death (SCD) observed in a subset of children with sensorineural deafness and otherwise unexplained heart rate–corrected QT interval prolongation on ECG stemmed from a novel congenital disorder.¹ This was followed by independent descriptions of a syndrome characterized by a cardiac-only phenotype consisting of prolonged QT intervals and an increased risk for syncope, seizures, and SCD in the absence of sensorineural deafness by Drs Cesarino Romano and Owen Ward, in 1963 and 1964, respectively.^2,3

During the past 5 decades, insights gleamed from a multitude of clinical, epidemiological, and molecular studies have demonstrated that LQTS is a collection of not only genetically and phenotypically diverse disorders of cardiac repolarization that encompasses the aforementioned predominantly autosomal-dominant, nonsyndromic Romano–Ward syndrome now simply referred to as LQTS⁴ and autosomal-recessive, multisystem, Jervell–Lange Nielsen syndrome but also exceedingly rare multisystem LQTS subtypes such as Timothy syndrome (TS) characterized by QT prolongation and an increased risk of SCD along with an array of extracardiac manifestations.

Although mutations in calcium (Ca²⁺)-handling proteins, including the CACNAlC-encoded L-type Ca²⁺ channel (LTCC), RYR2-encoded ryanodine receptor-2 intracellular calcium release channel (RyR2), and many other auxiliary interacting proteins, are central to the pathogenesis of inherited cardiac arrhythmia syndromes, such as catecholaminergic ventricular tachycardia and Brugada syndrome, until recently, the contribution of dysfunctional Ca²⁺ handling to the pathogenesis of LQTS was limited to the extremely rare and highly lethal multisystem TS. However, the unexpected recent discoveries of multiple nonsyndromic LQTS-causative mutations in CACNA1C,^5–7 syndromic LQTS-causative mutations in CALM1-3-encoded Calmodulin^8,9 and TRDN-encoded Triadin,¹⁰ and common genetic variation at several novel genetic loci that modulate QT interval duration in health¹¹ that encode or reside near known Ca²⁺-handling proteins suggest a larger role for Ca²⁺ cycling in cardiac repolarization and the pathogenesis of LQTS.

In this review, we examine existing paradigms and recent advances that shape our current understanding of the molecular basis of LQTS with a focus on the molecular insights provided by recent discoveries linking genetic variation in Ca²⁺-handling proteins to the pathogenesis of LQTS and modulation of cardiac repolarization in health.

Molecular Basis of Cardiac Ca²⁺ Cycling

The carefully regulated flux of Ca²⁺ within and through cardiomyocytes governs both cardiac excitability and contractileity. Not only does Ca²⁺ serve as the critical link between the electric stimuli generated by plasma membrane depolarization and mechanical contraction of the cardiomyocyte during excitation–contraction coupling (Figure 1)¹² but also it generates a substantial inward/depolarizing current that modulates action potential (AP) duration and regulates several intracellular processes, including gene transcription. Although a detailed discussion of the major molecular players in cardiac excitation–contraction coupling/Ca²⁺ cycling is outside the scope of this review, a thorough review of the genetics, structure, and functionality of the Ca_v1.2 LTCC and RyR2 macromolecular complexes is contained within the Data Supplement to provide interested readers with sufficient background to appreciate how perturbations in these molecular constituents may contribute to the pathogenesis of LQTS.

Figure 1.

Cardiac excitation-contraction coupling. Depolarization of the transverse tubule (T-tubule) activates voltage-gated L-type Ca²⁺ channels (LTCC, designated in schema as Cav1.2) allowing calcium to enter the cytosol (1). The small amount of calcium influx through the LTCC triggers the large-scale release of calcium from stores in the sarcoplasmic reticulum through ryanodine receptors (RyR2) within calcium release units (CRU) producing a cytosolic calcium transient in a process deemed calcium-induced calcium release (2). Increased cytosolic calcium concentrations activate the cardiac myofilaments by binding to troponin C thereby inducing allosteric changes in myosin that results in cardiac muscle contraction during systole (3). Intracellular calcium is pumped back into the SR by the SR Ca²⁺ ATPase (SERCA2a) and expelled from the cell via the Na+/Ca²⁺ exchanger (NCX) (4) enabling the relaxation of the myofilaments during diastole (5).

Molecular Basis of LQTS

During the early-to-mid 1990s, a series of linkage analysis and positional cloning studies identified KCNQ1 (LQT1), KCNH2 (LQT2), and SCN5A (LQT3) as genetic substrates for congenital LQTS.^13–18 Subsequent epidemiological studies ascertained that disease-causative mutations in these 3 major genes account for an estimated 60% to 75% of genotype-positive LQTS cases (Table I in the Data Supplement).^19,20

After the discovery of these 3 major LQTS-susceptibility genes, at least 14 additional minor LQTS-susceptibility genes have been described in the literature. The resulting LQTS subtypes can be further classified on the basis of whether they yield a nonsyndromic or multisystem clinical phenotype (Table I in the Data Supplement). Furthermore, because the majority of minor LQTS genes encode channel-interacting proteins that work in concert with the Na_v1.5, Ca_v1.2, K_v11.1, and K_v7.1 pore-forming α-subunits, a current-centric model has been proposed as a means of summarizing the genetic and electrophysiological basis of the major and minor LQTS-susceptibility genes (Figure 2).²¹

Figure 2.

Current-centric classification of long-QT syndrome (LQTS) genotypes. The clinical phenotypes resulting from the abnormal ventricular cardiac action potential depolarization (purple) or repolarization (orange) are grouped according to the specific current perturbed by an underlying genetic defect. Blue circles represent mutations that confer a loss of function to the specified current, whereas green circles confer a gain of function. Solid lines indicate those disorders that are autosomal dominant, whereas dashed lines indicate those disorders that are autosomal recessive. Solid black outlines indicate nonsyndromic genotypes and solid orange outlines represent multisystem genotypes. ABS indicates ankyrin-B syndrome; COTS, cardiac-only Timothy syndrome; JLNS, Jervell and Lange-Nielson syndrome; I_Ca,L, L-type calcium current; I_K1, inwardly rectifying potassium current; I_KAch, G-protein-coupled inwardly rectifying potassium current; I_Kr, rapid component of the delayed rectifier potassium current; I_Ks, slow-component of the delayed rectifier potassium current; I_Na, cardiac sodium current; and TKO, triadin knockout syndrome. Adapted with permission from Giudicessi and Ackerman²¹. Authorization for this adaptation has been obtained both from the owner of the copyright in the original work and from the owner of copyright in the translation or adaptation.

Because the molecular basis of the major and minor LQTS-susceptibility genes linked to the function of I_Na, I_Kr, and I_Ks currents have been described recently in detail elsewhere,²² the following sections focus on examining the ≈5% to 10% of multisystem and nonsyndromic LQTS cases that arise from Ca²⁺-handling protein dysfunction.

TS: Aberrant Ca²⁺-Dependent and Ca²⁺-Independent Intracellular Processes

TS is an extremely rare, largely sporadic form of multisystem LQTS characterized by severe QT prolongation on ECG (average heart rate–corrected QT interval ≈580 ms), high incidence of life-threatening cardiac arrhythmias/SCD, syndictyly, minor craniomaxillofacial abnormalities, and baldness at birth and variably expressed intermittent hypoglycemia, immunodeficiency, congenital heart defects, and developmental delay/neurocognitive impairments along the autism spectrum.^23–25

Initially, TS was thought to arise from a single recurrent mutation, G406R, in the alternatively spliced CACNA1C exon 8a that accounts for ≈20% of the Ca_v1.2 mRNA transcripts in the heart and brain and encodes the distal portion of the domain I transmembrane segment 6 (IS6) known to play a critical role in the voltage-dependent inactivation (VDI) of Ca_v1.2 (Figure 3).^25–27 However, shortly after the elucidation of the molecular basis of typical TS now referred to as type I TS (TS1), 2 cases of atypical or type II TS (TS2) were reported in the literature.²⁸ Interestingly, both cases were characterized by extreme QT prolongation (average ≈640 ms) complicated by multiple episodes of cardiac arrest in the absence of syndictyly and harbored de novo mutations in the CACNA1C exon 8–containing heart- and brain-predominant Ca_v 1.2 splice variant.²⁸ Interestingly, the exon 8 CACNA1C-G406R–positive TS2 proband displayed severe neurodevelopmental delay, craniofacial/odontic abnormalities, and possible nemaline skeletal myopathy during the first few months of life, whereas the exon 8 CACNA1C-G402S–positive TS2 proband seemed developmentally normal until suffering an out-of-hospital cardiac arrest at age 4.²⁸ It was postulated initially that the lack of syndactyly and the potentially more severe neurological/cardiac phenotypes anecdotally observed in TS2 versus TS1 were likely secondary to the differential tissue-specific expression of exon 8– and exon 8a–containing Ca_v 1.2 splice variants.²⁸ Although this is likely the case for exon 8 G406R–mediated TS2, the recent identification of exon 8 G402S in an otherwise developmentally normal adolescent woman with borderline QT prolongation who presented with an out-of-hospital cardiac arrest²⁹ coupled with the normal development of the initial G402S-positive TS2 proband before his first cardiac arrest²⁸ suggest that (1) CACNA1C-G402S results in a milder clinical phenotype more akin to nonsyndromic LQTS and (2) the neurodevelopmental sequelae observed in the initial CACNA1C-G402S–positive TS2 proband are more likely secondary to arrhythmia-induced anoxic brain injury than underlying CACNAlC-mediated developmental delay.

Figure 3.

Electrophysiological and clinical manifestations of putative long-QT syndrome (LQTS)-causative gain-of-function CACNA1C mutations. A, The location of Timothy syndrome (TS; maroon circles)-, cardiac-only Timothy syndrome (COTS; orange circles)-, and LQTS (CACNA1C-LQTS; green circles)-causative CACNA1C mutations are depicted on the Ca_y1.2 linear protein topology. Biogenic (small dash), kinetic (solid), and mixed (large dash) generalized electrophysiological manifestations of individual CACNAlC mutations are indicated by contrasting circle outlines. B, Summary of the electrophysiological and clinical manifestations of TS, COTS, and LQTS. EP indicates electrophysiological; HCM, hypertrophic cardiomyopathy; and SCD, sudden cardiac death.

Mechanistically, the IS6 and α-interaction domain–containing I-II loop linker of the Ca_v1.2 α-subunit form a continuous α-helix–rich region that allow Ca_vβ auxiliary subunits to bind and modulate Ca_v 1.2 gating in a manner that supports faster VDI.^30,31 In addition, calmodulin (CaM)-mediated Ca²⁺-dependent inactivation (CDI) may also be dependent on Ca_vβ and the structural integrity of the Ca_v1.2 α-subunit IS6-α-interaction domain region suggesting that the Ca_v1.2 IS6-α-interaction domain/Ca_vβ complex serves as the common denominator by which VDI and CDI modulate Ca_v 1.2 gating.³² Given the unique conformational flexibility imparted by glycine (G) residues and the common localization of TS-associated mutations to the distal IS6 region of Ca_v 1.2, it is therefore not surprising that both TS1- and TS2-associated CACNA1C exon 8a (G406R) and exon 8 (G402S and G406R) mutations drastically reduce Ca_v 1.2 VDI in vitro possibly via the alteration of Ca_vβ binding/function.^25,28

Although subsequent studies, including those in TS-specific induced pluripotent stem cell (iPSC)–derived cardiomyocytes³³ and cortical neurons,³⁴ have reaffirmed the effect of CACNA1C-G406R on VDI, the effects of CACNA1C-G406R on CDI reported to date are widely discordant.^28,35,36 Furthermore, several alternatives to the IS6-α-interaction domain/Ca_vβ-mediated reduction in VDI hypothesis have been proposed, including formation of aberrant A kinase anchor protein–Ca_v1.2 complexes³⁷ and facilitation by aberrant/excess calmodulin kinase II activity.^38,39As such, the precise mechanisms by which TS1-and TS2-causative mutations reduce Ca_v1.2 VDI and the potential contribution of impaired Ca_v 1.2 CDI to the generation of the multisystem TS phenotype remain incompletely understood.

In the heart, the reduction of Ca_v 1.2 VDI conferred by TS1-and TS2-causative CACNA1C mutations is predicted in silico to accentuate the inward depolarizing current and prolong AP repolarization/duration.^25,28 Although subsequent studies in both iPSC cardiomyocytes³³ and a TS mouse model⁴⁰ concluded that increased Ca²⁺ influx through mutant Ca_v1.2-TS channels during the plateau phase prolongs cardiac AP duration, both studies also demonstrated that TS ventricular myocytes are in a constant state of Ca²⁺ overload and thus prone to spontaneous ectopic Ca²⁺ release from sarcoplasmic reticulum (SR) and the generation of potentially arrhythmogenic delayed after-depolarizations (DADs). As such, the proarrhythmic state observed in TS likely arises from the generation of early after-depolarizations secondary to increased myocyte refractoriness and DADs secondary to spontaneous SR Ca²⁺ release during phase 2/3 and phase 4 of the cardiac AP, respectively. Interestingly, roscovitine (Seliciclib), a cyclin-dependent kinase inhibitor previously shown to enhance Ca_v1.2 VDI in heterologous expression systems via direct extracellular binding,^41,42 rescued the electrophysiological perturbations observed in TS iPSC-derived cardiomyocytes,³³ suggesting that agents with similar mechanisms of action may prove beneficial in the treatment of TS.

Although the contribution of increased Ca²⁺ influx to the proarrhythmic TS cardiac phenotype is relatively well understood, precisely how TS-causative CACNA1C mutations generate a myriad of variably expressed extracardiac phenotypes, particularly in nonexcitable tissues, remains unclear. At present, there is evidence to suggest that Ca²⁺-dependent processes likely underlie the intermittent hypoglycemia (excessive Ca²⁺-mediated insulin release from pancreatic β cells)⁴³ and craniofacial dysmorphism (mandibular chondrocyte/osteoblast hypertrophy via increased Ca²⁺-dependent calcineurin/nuclear factor of activated T-cell transcription factors signaling)⁴⁴ aspects of the multisystem TS phenotype. However, the apparent lack of neurodevelopmental delay and other extracardiac manifestations in the 2 CACNA1C-G402S–positive TS2 cases,^28,29 despite the fact that CACNA1C-G402S and CACNA1C-G406R result in similar increases in Ca²⁺ influx in vitro, suggests that perturbation of Ca²⁺-independent processes may also contribute to multiorgan dysfunction in TS.

Consistent with this theory, Krey et al⁴⁵ demonstrated that TS1 iPSC-derived neurons displayed marked impairment of dendrite formation, a common feature of neurodevelopmental/autism spectrum disorders, independent of Ca²⁺ influx through mutant TS1-Ca_v1.2 channels. Based on an elegant series of experiments, Krey et al⁴⁵ propose that conformational changes in TS1-Ca_v 1.2 channels, unrelated to their Ca²⁺-permeating properties, lead to decreased binding/recruitment of Gem, a guanosine triphosphate-binding protein, to the Ca_v 1.2 macromolecular complex which then precipitates the loss of RhoA GTPase inhibition, aberrant cytoskeletal remodeling, and ultimately dendritic retraction in TS1 iPSC-derived neurons. As such, there is ample evidence to suggest that electrophysiological perturbations in L-type calcium current (I_CaL)^33,40 underlie the proarrhythmic component of the TS cardiac phenotype, whereas disruption of downstream Ca²⁺-dependent (eg, calcineurin/nuclear factor of activated T cells)⁴⁴ and Ca²⁺-independent (eg, Gem/RhoA/ROCK)⁴⁵ signaling cascades may underlie many of the extracardiac manifestations.

Lastly, our current understanding of the molecular basis of TS is further complicated by the recent identification of 3 additional CACNA1C mutations (CACNA1C-G402R, -I1166T, and -A1473G), including 2 mutations that reside outside of exon 8/8a, in patients with TS-like phenotypes (Figure 3).^46–48 Interestingly, all putative TS-causative mutations described to date localize to the hydrophobic residuerich boundaries between the last transmembrane segment (S6) and interdomain linkers of their respective Ca_v1.2 domains/repeats (Figure 3), suggesting that these regions may contribute to a key tertiary/quaternary structure(s) within the Ca_v 1.2 channel that regulate gating kinetics and function as a scaffold for the interaction/binding of downstream signaling proteins. At present, it remains to be seen if the novel phenotypes (eg, hypotonia, joint hyperflexibility/contractures, clinodactyly, etc)^46,47 associated with these recently described TS-like cases represent true genotype-phenotype correlations secondary to unique electrophysiological perturbations (eg, increased I_CaL window current with CACNA1C-I1166T)⁴⁷ or are simply the result of variable expressivity. Regardless, these findings suggest that expanded CACNA1C sequencing should be considered for any patient with a TS-like phenotype and previous negative exon 8/8a targeted screening.

Nonsyndromic Long-QT Syndrome: Evolving Role of the L-Type Ca²⁺ Channel

In 2013, a combination of whole-exome sequencing, gene prioritization, and candidate-based screening, unearthed 4 novel putative disease-causative missense mutations in CACNA1C in LQTS patients with isolated QT prolongation devoid of any congenital heart defects or extracardiac manifestations that define TS clinically.⁵ Interestingly, 3 out of 4 LQTS-causative CACNA1C mutations (K834E, P857L, and P857R) localized to the PEST (proline [P]-, glutamic acid [E]-, serine [S]-, and threonine [T]-rich) domain of Ca_v1.2’s II-III linker (Figure IA in the Data Supplement) that is believed to serve as a proteolytic signal peptide that when unmasked marks the LTCC for rapid degradation via calpain-mediated and ubiquitin-proteasome-mediated mechanisms.^5,49–51

Given the previous observation that deletion of the Ca_v1.2 LTCC’s 2 PEST domains (PEST1 in the I-II linker and the aforementioned PEST3 in the II-III linker) individually increased the stability and current density of heterologously expressed Ca_v1.2 channels,⁵⁰ it is not entirely surprising that in vitro functional characterization of P857R-CACNA1C revealed a significant increase in both Ca_v1.2 cell surface expression and peak current density without significantly altering Ca_v1.2 channel kinetics.⁵ Collectively, these findings suggest that rare CACNA1C mutations localizing to the Ca_v1.2 PEST3 domain may disrupt normal PEST3-mediated Ca_v1.2 degradation and impart a biogenic I_CaL gain of function that manifests clinically as an isolated LQTS phenoltype without any other cardiac or extracardiac abnormalities (CACNA1C-LQTS).

After the initial whole-exome sequencing–aided discovery by Boczek et al,⁵ screening of CACNA1C in 2 independent LQTS cohorts identified 8 additional LQTS-causative mutations shown to impart an I_CaL gain of function in vitro through a variety of electrophysiological mechanisms (Figure 3).^6,7 In addition, subsequent screening of genotypenegative LQTS individuals with documented echocardiographic evidence of hypertrophic cardiomyopathy and a family history of hypertrophic cardiomyopathy–like phenoltypes led to the identification of a novel genetic hotspot in the distal portion of the Ca_v1.2 I-II linker (CACNA1C exon 12) responsible for a newly described clinical entity, cardiac-only TS (COTS), characterized by the concomitant but variably expressed phenotypes of LQTS, hypertrophic cardiomyopathy, congenital heart defects, and SCD in the absence of any extracardiac symptoms (Figure 3).⁵² With these discoveries, the initial hypothesis that TS and CACNA1C-LQTS may be differentiated by distinct electrophysiological mechanisms was turned on its head as several of the newly discovered CACNA1C-LQTS/COTS mutations imparted an I_CaL gain of function via the slowing of VDI analogous to the effect observed for the TS1- and TS2-causative CACNA1C mutations.^6,7 Furthermore, in silico action potential modeling predicted that many CACNA1C-LQTS mutations result in a more pronounced I_CaL gain of function (AP duration prolongation and increased cytosolic Ca²⁺ concentration) than the canonical TS exon 8a G406R-CACNA1C mutation when the estimated percentage of affected Ca_v1.2 transcripts are accounted for.⁷ These findings suggest that the stark phenotypic differences between TS, COTS, and CACNA1C-LQTS cannot be explained by distinct electrophysiological mechanisms readily appreciated in heterologous expression systems, the degree of increased cytosolic Ca²⁺ influx, or mutation localization to specific Ca_v1.2 domains such as the S6/interdomain linker boundaries as summarized in Figure 3. As such, the extracardiac manifestations of TS syndrome and the cardiac hypertrophy/congenital heart defects observed in TS and COTS may be (1) independent of the Ca²⁺-permeating properties of mutant Ca_v1.2 channels (eg, perturbation of aforementioned Ca²⁺-dependent (calcineurin/nuclear factor of activated T cells) or Ca²⁺-independent (GEM/RhoA/ROCK) signaling cascades that culminate in aberrant gene expression) and (2) modulated by as of yet unrecognized genetic/epigenetic mechanisms (eg, interindividual variation in splicing patterns, auxiliary subunit expression/function, or regulatory noncoding RNAs).⁵³

Calmodulinopathic LQTS: Impaired LTCC Ca²⁺-Dependent Inactivation

Over the past several years, independent whole-exome sequencing studies by Crotti et al⁸, Makita et al⁵⁴, and Reed et al⁹ linked heterozygous sporadic/de novo mutations in the identical CALM1-, CALM2-, and CALM3- encoded CaM, a ubiquitous Ca²⁺ sensor essential for the Ca²⁺-dependent regulation of an array of intracellular processes, to what was described initially as a rare multisystem disorder characterized by marked QT prolongation and recurrent cardiac arrest (often in infancy) accompanied by variably expressed congenital heart defects, seizures, and neurodevelopmental delays. However, the recent descripttion of CALM2-positive cases with a milder phenotype consisting of later onset cardiac events (>1 year of life) in the absence of discernible neurodevelopmental delay,⁵⁴ raises the possibility that the phenotype initially observed in calmodulinopathic LQTS⁸ may (1) depend on the severity of the Ca²⁺ binding impairment/reduction in Ca²⁺ affinity conferred by mutant CaMs, (2) occur secondary to the poorly understood effects of the differential temporospatial/tissue-specific expression of the 3 CALM genes (in the human heart CALM1, CALM2, and CALM3 is expressed at ≈1:2:5 ratio),⁸ and (3) represent neurological sequelae of recurrent cardiac arrests early in life rather than an intrinsic manifestation of CaM mutations.

Interestingly, all LQTS-causative CALM1–3 mutations described to date localize at or immediately proximal to Ca²⁺⁻ coordinating residues of the C-terminal lobe (C-lobe) of CaM (Figure 4A) and impart a marked reduction in Ca²⁺-binding affinity.^8,9,54 Despite the fact that CaM regulates a number of cardiac ion channels, the predominant effect of LQTScausative CALM1-3 mutations seems to be the loss of CDI resulting in unrestrained Ca²⁺ influx as the effects on RyR2 (reduced CaM binding) and Na_v1.5 (increased I_Na late current) were mild and widely variable.^55–58 In contrast, CaM mutations that produce a predominantly catecholaminergic ventricular tachycardia phenotype localize to both the N-terminal-lobe (N-lobe; N54I) and C-lobe (N98S) and increase RyR2-binding affinity/single channel open probability and the incidence of spontaneous Ca²⁺ release from the SR with little to no effect on Ca²⁺ binding affinity (Figure 4A).^57,58

Figure 4.

Molecular basis of calmodulinopathic long-QT syndrome (LQTS). A, The location of LQTS (maroon solid circles), catecholaminergic ventricular tachycardia (CPVT; green solid circles), and LQTS/CPVT overlap (green/maroon dashed circles)-causative CALM1, CALM2, and CALM3 mutations in the identical calmodulin linear protein topology. B, Schematic model for dual-phase L-type calcium channel (LTCC) calcium-dependent inactivation (CDI). The amino (N)-and carboxy (C)-terminal lobes of calcium-free calmodulin (apoCaM) preassociate with the LTCC carboxy (C)-terminal IQ motif-containing calcium inactivation region. After activation of the LTCC, binding of calcium to calmodulin (Ca²⁺/CaM) induces a conformational shift that allows the N-lobe and C-lobe to bind the N-terminal spatial Ca²⁺-transforming element (NSCaTE) and C-terminal IQ motif, respectively, leading to CDI. C, Schematic model for perturbed LTCC dual-phase calciumdependent inactivation in calmodulinopathic LQTS. LQTS-causative CALM mutations result in loss of calcium-binding affinity to the C-lobe of calmodulin and reduce/abolish the rapid component of CDI leading to intracellular calcium overload.

The fact that all LQTS-causative CaM mutations, including those that yield a LQTS/catecholaminergic ventricular tachycardia overlap phenotype (N98S, D132E, and Q136V), localize to the CaM C-lobe raises the possibility of a functional bipartition whereby the CaM N-and C-lobes are responsible for different sets of cellular functions. One example of partitioned CaM function potentially pertinent to the pathogenesis of calmodulinopathic LQTS is the presence of a dual-phase CDI unique to LTCCs (Figure 4B). In comparison to other voltage-gated calcium channel classes where the lower Ca²⁺ affinity N-lobe mediates CDI and the higher Ca²⁺ affinity C-lobe either potentiates Ca²⁺ entry during Ca²⁺-dependent facilitation or plays no role in Ca²⁺-dependent regulation,^59,60 the LTCCs’ N-lobe and C-lobe underlie distinct slow and rapid components of CDI, respectively (Figure 4B).⁶¹ As such, the localization and function of LQTS-causative CALM mutations as well as modest effects on RyR2 or Na_v1.5 suggest that calmodulinopathic LQTS primarily arises from the impairment of C-lobe–mediated rapid CDI (Figure 4C).

However, given the ubiquitous nature of CaM, significant functional redundancy, and vast number of intracellular CaM targets, without more in-depth studies, it remains to be seen if more pervasive defects in ion channel regulation (eg, K⁺ channels, other voltagegated Ca²⁺ channels, etc) and CaM-dependent signaling cascades contribute to the molecular basis of calmodulinopathic LQTS.

Triadin Knockout Syndrome: Remodeling of the Calcium Release Unit Molecular Architecture

Most recently, Altmann et al¹⁰ described a rare autosomalrecessive form of LQTS characterized by transient/consistent QT prolongation with extensive precordial (V₁ through V₄) T-wave inversions, severe and often β-blocker- and left cardiac sympathetic denervation-refractory exercise-induced cardiac events (eg, syncope, sudden cardiac arrest, and SCD) in early childhood, and mild-to-moderate proximal skeletal muscle weakness that arises secondary to either homozygous or compound heterozygous frameshift/null mutations in TRDN-encoded triadin, a key structural component of the cardiac release unit (Figure 5A).

Figure 5.

Molecular basis of triadin knockout (TKO) syndrome. A, Artistic rendering of the molecular architecture of a normally functioning cardiac dyad composed of closely juxtaposed Ca_v1.2 L-type calcium channels (LTCC) and ryanodine receptor-2 (RyR2) intracellular calcium release channels and appropriate sequestration of calcium in the sarcoplasmic reticulum (SR) as a result of the appropriate expression/function of triadin, junctin, junctophilin-2, and calsequestrin in calcium release units (CRUs). B, Artistic rendering of the remodeled cardiac dyad molecular architecture in TKO syndrome and the resulting Ca²⁺ overload that occurs secondary to the loss of Ca_v 1.2 LTCC and RyR2 juxtaposition, normal triadin, junctin, and junctophilin expression/function, and proper calcium/calsequestrin localization in the SR based on observations in triadin-null mice.

Although no functional studies were undertaken with the initial clinical and genetic description of Triadin knockout (TKO) syndrome, insights gleamed from ventricular arrhythmia-prone TRDN-null mice suggest that complete ablation of triadin, as would be expected in TKO syndrome patients, significantly (1) disrupts the approximation of the T-tubule and the junctional SR within the cardiac dyad, (2) reduces the expression of key junctional SR proteins including RyR2, calsequestrin 2, and junctin, and (3) reduces the colocalization/juxtaposition of Ca_v1.2/RyR2 and RyR2/Calsequestrin2 in the cardiac release unit (Figure 5B).⁶² This radical remodeling of the cardiac release unit molecular architecture reduces SR Ca²⁺ release leading to impaired Ca_v1.2 LTCC CDI, unrestrained cytosolic Ca²⁺ influx via I_CaL, and subsequent SR Ca²⁺ overload (Figure 5B).^62–64 Interestingly, in TRDN-null myocytes, Ca_v1.2 activation and inactivation remain slower even when Ba²⁺, which does not trigger SR Ca²⁺ release or Ca_v1.2 CDI, is used as the charge carrier, suggesting that the loss of Triadin intrinsically alters Ca_v1.2 gating by as of yet undefined mechanisms independent of SR Ca²⁺-release–triggered Ca_v1.2 CDI_.^62,65

Collectively, the misappropriation of calsequestrin 2 in the SR,^62,66 SR Ca²⁺ overload,⁶² and reduction in junctin-mediated RyR2 inhibition at high SR luminal Ca²⁺ concentrations⁶⁷ that arise secondary to remodeling of cardiac release unit molecular architecture in TRDN-null mice is hypothesized to increase the frequency of spontaneous ectopic Ca²⁺ release from the SR leading to proarrhythmic DADs, particularly in the setting of β-adrenergic stimulation.^62,65 In addition, the slowed Ca_v1.2 inactivation observed in TRDN-null mice theoretically could prolong action potential duration and create an electric substrate favorable for the generation of early after-depolarization–triggered ventricular arrhythmias, the predominant arrhythmogenic mechanism in most LQTS subtypes. Thus, in TKO syndrome, calmodulinopathic LQTS, and to a lesser extent in non-Ca²⁺-mediated LQTS subtypes (eg, I_Na-mediated LQT3) the perturbation/modulation of a myriad of Ca²⁺-dependent events, including disrupted Ca_v1.2 kinetics/enhanced I_Ca,L current, increased spontaneous SR Ca²⁺ release via RyR2, and promotion of SR Ca²⁺ loading via NCX, likely contribute to an underlying proarrhythmic electrophysiological substrate capable of triggering DAD- and early after-depolarization–mediated arrhythmias.^55,56,62,68 Although it is not clear whether a DAD- or early afterdepolarization–mediated mechanism is predominantly responsible for the ventricular arrhythmias observed in TRDN-null mice, treatment of TRDN-null myocytes with the DHP Ca²⁺ channel blocker nifedipine abolished SR Ca²⁺ overload and spontaneous SR Ca²⁺ release, suggesting that DHPs and other class IV antiarrhythmics may one day play a role in the treatment of patients with TKO syndrome.^10,62

Ultimately, these observations in TRDN-null mice function to substantiate the hypothesized molecular basis of TKO syndrome in humans and provide a reasonable launching point for future inquiries aimed at elucidating the precise arrhythmogenic mechanism(s) that underlie the severe medically refractory adrenergically mediated cardiac events that plague TKO syndrome patients with the hope that these insights will catalyze novel therapeutic strategies.

Modulation of Cardiac Repolarization by Ca²⁺ Handling Proteins in Health

In recent years, the association of QT prolongation/shortening with an increased risk of SCD in the general population^69–71 and the barrier that drug-induced QT prolongation has presented to new drug development has served as an impetus for QT interval genome-wide association studies in several large population-based cohorts such as the Framingham Heart Study and the Cardiovascular Heart Study. Beginning in 2009, 3 large genome-wide association study meta-analyses, including the mammoth QT-IGC consortium expanded genome-wide association study meta-analysis of 76 061 individuals of European descent,¹¹ built on existing single cohort studies by collectively demonstrating that common variants in 6 established (NOS1AP, KCNQ1, KCNE1, KCNH2, SCN5A, and KCNJ2) and 29 novel genetic loci collectively explain ≈8% to 10% of heritable QT interval variability.^11,72,73

Interestingly, a surprising number of these novel genetic loci localize within or near genes that encode intracellular Ca²⁺-handling proteins, including ATP2A2-encoded SR ATPase 2a calcium pump, SERCA-regulator PLN-encoded phospholamban, and the SLC8A1-encoded NCX1 Na⁺/Ca²⁺ exchanger whose roles in excitation–contraction coupling were discussed previously and are summarized in Figure 1.¹¹ Furthermore, mutational analysis of the 6 genes (ATPA2A2, CAV1, CAV2, SLC8A1, SRL, and TRPM7), in closest proximity to the novel genetic loci with the strongest statistical signifycance and highest biological plausibility in an international cohort of 298 unrelated LQT1-LQT3 negative LQTS probands, identified potentially pathogenic frameshift mutations in ATP2A2-encoded SR ATPase 2a and the TRPM7-encoded transient receptor channel melastatin 7 Mg²⁺/Ca²⁺ channel known to mediate Ca²⁺ influx and regulate expression of other Ca²⁺-handling proteins.¹¹ Although much work remains to fully understand the physiological ramifications of common and rare genetic variation in these newly identified genetic loci, the above studies highlight a larger than anticipated role for Ca²⁺ flux in cardiac repolarization/modulation of QT interval in health and hint at a potentially even greater role for Ca²⁺-handling proteins in the pathogenesis of congenital and acquired LQTS.

Concluding Remarks

The discovery and subsequent investigation of rare and common genetic variation in Ca²⁺-handling proteins has illuminated a previously underappreciated and rapidly expanding role for Ca²⁺ signaling in cardiac repolarization and the pathogenesis of nonsyndromic and multisystem forms of congenital LQTS. Undoubtedly, these discoveries will provide a rich foundation for future inquiries, aided by the continued use of next-generation sequencing technologies and patient-specific iPSC-derived cardiomyocytes, aimed at (1) determining the genetic substrates that underlie the roughly 15% to 20% of LQTS cases that remain genetically elusive, (2) elucidating the novel genetic/epigenetic determinants that underlie the variable phenotypic expressivity observed in most LQTS subtypes and the dramatic phenotypic divergences observed between CACNA1C-mediated TS, CACNA1C-mediated COTS, and CACNA1C-LQTS, (3) developing a deeper understanding of the pathophysiological mechanisms of LQTS-causative mutations in Ca²⁺-handling proteins in hopes that such insights will lead to the development of individualized, genotype-guided therapeutic strategies with improved efficacy and safety profiles, and (4) further investigating the potential contribution of an increased burden of common genetic variation in Ca²⁺-handling proteins to the pathogenesis of drug-induced LQTS. Because we continue to revisit the role of Ca²⁺ in cardiac repolarization and the pathogenesis of nonsyndromic and multisystem forms of LQTS, there is renewed hope that insights gleamed from these efforts will lead to the development of novel approaches to the diagnosis, risk stratification, and treatment of LQTS patients, particularly those with high-risk LQTS subtypes such as TS, calmodulinopathic LQTS, and TKO syndrome who are among those at greatest risk of succumbing to potentially lethal ventricular arrhythmias.