Reduction in Junctophilin 2 Expression in Cardiac Nodal Tissue Results in Intracellular Calcium-driven Increase in Nodal Cell Automaticity

Abstract

Background:

Spontaneously depolarizing nodal cells comprise the pacemaker of the heart. Intracellular calcium (Ca²⁺) plays a critical role in mediating nodal cell automaticity and understanding this so-called “Ca²⁺ clock” is critical to understanding nodal arrhythmias. We previously demonstrated a role for junctophilin-2 (Jph2) in regulating Ca²⁺-signaling through inhibition of ryanodine receptor 2 (RyR2) Ca²⁺ leak in cardiac myocytes; however, its role in pacemaker function and nodal arrhythmias remains unknown. We sought to determine whether nodal Jph2 expression silencing causes increased sinoatrial and atrioventricular nodal cell automaticity due to aberrant RyR2 Ca²⁺ leak.

Methods:

A tamoxifen-inducible, nodal tissue-specific, knockdown mouse of Jph2 was achieved using a Cre recombinase-triggered short RNA hairpin directed against Jph2 (Hcn4:shJph2). In vivo cardiac rhythm was monitored by surface ECG, implantable cardiac telemetry, and intracardiac electrophysiology (EP) studies. Intracellular Ca²⁺ imaging was performed using confocal-based line scans of isolated nodal cells loaded with fluorescent Ca²⁺ reporter Cal-520. Whole cell patch clamp was conducted on isolated nodal cells to determine action potential kinetics and Ncx1 function.

Results:

Hcn4:shJph2 mice demonstrated a 40% reduction in nodal Jph2 expression, resting sinus tachycardia and impaired heart rate response to pharmacologic stress. In vivo intracardiac EP studies and ex vivo optical mapping demonstrated accelerated junctional rhythm originating from the atrioventricular node. Hcn4:shJph2 nodal cells demonstrated increased and irregular Ca²⁺ transient generation with increased Ca²⁺ spark frequency and Ca²⁺ leak from the sarcoplasmic reticulum. This was associated with increased nodal cell AP firing rate, faster diastolic repolarization rate, and reduced Ncx1 activity during repolarized states compared to control. Phenome-wide association studies (PheWAS) of the JPH2 locus identified an association with sinoatrial nodal disease and atrioventricular nodal block.

Conclusions:

Nodal-specific Jph2 knockdown causes increased nodal automaticity through increased Ca²⁺ leak from intracellular stores. Dysregulated intracellular Ca²⁺ underlies nodal arrhythmogenesis in this mouse model.

Introduction

The nodal and conduction system of the heart play a critical role in the generation and propagation of the electrical depolarization which initiates cardiac myocyte contraction¹. In the normal heart, the sinoatrial (SA) node automatically depolarizes and triggers a wave of electrical depolarization that causes atrial contraction^{2, 3}. Depolarization is then delayed at the atrioventricular (AV) node, allowing for contraction of the atria and ventricular filling, before being propagated down ventricular bundle branches and the Purkinje fiber network. This coordinated depolarization of the ventricles allows for efficient contraction that ejects blood into pulmonary and systemic circulation. A broad class of arrhythmias, nodal arrhythmias are characterized by impaired automaticity or function of the SA and AV nodes, and are associated with fatigue, syncope, and even death⁴. Arrhythmias impacting the nodal or electrical conduction system of the heart are common and impact up to ~1% of the population, yet there are limited pharmacologic therapies to treat these arrhythmias⁵. This reflects the need for additional understanding of the molecular mechanisms of nodal cell automaticity.

Classically, spontaneous action potential generation of pacemaker cells has been viewed as a coordinated event among membrane-limited ion channels creating a so-called “membrane voltage clock”⁴. Recent evidence from multiple investigators suggests that there is an intracellular “calcium (Ca²⁺) clock” which plays a critical role in impulse generation through rhythmic, spontaneous release of Ca²⁺ from the intracellular sarcoplasmic reticulum (SR) store via RyR2⁶. Indeed, nodal cell Ca²⁺ signaling may be a potential molecular target for pharmacologic manipulation as altered Ca²⁺ signaling has been associated with both physiological nodal cell depolarization and dysfunction^{7, 8}. Jph2-encoded junctophilin-2 (Jph2) is a structural protein that plays a critical role in Ca²⁺ regulation within the contractile myocytes of the heart⁹. We have previously shown that Jph2 expression knockdown in cardiomyocytes causes increased Ca²⁺ leak from intracellular stores (i.e. the SR) resulting in heart failure through dysregulation of the intracellular Ca²⁺ release channel RyR2^10-12. Follow-up studies have shown that Jph2 serves as a molecular regulator of intracellular Ca²⁺ and that a defective Jph2-RyR2 interaction may be targeted by small molecules to prevent pathologic Ca²⁺ leak¹³. This makes Jph2 a unique molecule to dissect the underlying role of Ca²⁺ dysregulation in nodal disease and a possible target for therapeutic intervention.

To this end, we report a conditional, nodal-specific Jph2 knockdown which causes increase nodal cell automaticity. Further, we found that this increase in automaticity is associated with increased SR store Ca²⁺ leak. Finally, phenome-wide association studies (PheWAS) of the JPH2 locus identified an association with sinoatrial nodal disease and atrioventricular nodal block. These findings suggest a possible role for Jph2 deficit-induced aberrant Ca²⁺ homeostasis and signaling in the development of nodal arrhythmias.

Methods

The data the support the findings of this study are available from the corresponding author upon reasonable request

Transgenic Mice

All mice were treated in accordance with the Baylor College of Medicine and the Duke University School of Medicine Institutional Animal Care and Use Committees, respectively, and reflected proposed guidelines for rigor and reproducibility in preclinical animal studies¹⁴. A nodal-specific, conditional Jph2 knockdown transgenic line was achieved by crossing two lines to create a double transgenic line. The nodal-specific Hcn4-KiT Cre mouse was previously described¹⁵. Briefly, this line expresses tamoxifen-inducible CreER^T2 inserted into the pacemaker channel Hcn4 locus (Hcn4). Previous studies have demonstrated robust expression of Cre selectively to nodal tissue without modulation of cardiac rhythm, basal heart rate (HR), or HR variability¹⁵. This line was crossed with a mouse line expressing a Jph2 shRNA¹². This shJph2 was placed under the transcriptional control of a high-expression U6 promoter inhibited by a neomycin cassette with flanking loxP sites. Presence of Cre-recombinase floxed out this inhibitory cassette allowing for inducible shJph2 expression. This created the Hcn4:shJph2 double transgenic line which was the experimental group (referred to as shJph2). Mice were injected IP with tamoxifen at approximately 3 months of age as described¹², and all experiments were conducted at 4-6 months of age. Littermate control mice carrying the Hcn4 Cre (Hcn4:WT) following tamoxifen injection were used as the controls for all studies. Cre-expression was confirmed utilizing the mTmG double-fluorescent Cre reporter mouse that expresses membrane-targeted tdTomato (mT) prior to Cre-recombination, and membrane-targeted GFP (mG) post Cre-excision¹⁶. All transgenic mice were on a C57BL/6J background. Due to sex-specific effects of tamoxifen on female mice physiology, including the cardiovascular system, male mice were utilized^{17, 18}. Mice were euthanized under isoflurane anesthesia using cervical dislocation. Additional details, including additional methods, are included in the Supplemental Materials.

Results

Hcn4:shJph2 mice demonstrate nodal-specific Jph2 knockdown

To determine whether loss of Jph2 expression results in increased nodal automaticity, a double transgenic mouse line Hcn4:shJph2 was created (Figure 1A). Following tamoxifen induction (Figure 1B), robust Cre-expression was confirmed by crossing the Hcn4:shJph2 mice with the mTmG reporter line which shifts changes RFP expression into GFP in the presence of Cre. We identified robust GFP expression in the SA node, AV node, and His-Purkinje/bundle branches, in cardiac sections from Hcn4:mTmG mice (Sup Figure I). Hcn4:shJph2 mice demonstrated normal growth and survival both pre- and post-induction with no loss of cardiac systolic function by transthoracic echocardiogram. Further, Hcn4:shJph2 mice demonstrated a 42.5±0.4% reduction in Jph2 mRNA expression by qPCR in excised AV nodal tissue compared to control (P<0.0001, Figure 1C). Further, immunofluorescence of Jph2 in Hcn4-expressing cells identified by histology of Hcn4:shJph2:mTmG mouse hearts identified a 33±2.5% reduction in Jph2 fluorescent intensity compared to control Hcn4:WT:mTmG control mice in both the SA and AV node (P<0.0001, Sup Figure II). Hcn4 mRNA levels were not altered in Hcn4:shJph2 SA nodal tissue compared to Hcn4:WT controls (Figure 1D). Further, there was no change in mRNA expression of Ryr2 and the cardiac sodium-calcium exchanger (Ncx1) in knockdown mice compared to controls (Figure 1E and 1F). To exclude the possibility that any heart rate or rhythm abnormality was secondary to elevated plasma catecholamine levels, serum ELISA was performed which demonstrated no alteration to circulating epinephrine and norepinephrine levels in Hcn4:shJph2 mice compared to controls (Figure 1G).

Figure 1:

Experimental mouse Hcn4:shJph2. A) A schematic demonstrating the tamoxifen-inducible Cre-recombinase with expression limited to Hcn4-expressing nodal cells (Hcn4) crossed with an shJph2 expression cassette under a Neo-inhibited promoter (shJph2) to make the experimental line (Hcn4:shJph2). Cre removes an inhibitory neo cassette which allows for loxP excision of the Neo-cassette and fusion of the pU6 promoter to drive expression of an anti-Jph2 short hairpin. B) A schematic of the timeline for induction of Cre by tamoxifen injection and experimental age. C-F) A bar graph demonstrating Jph2, HCN4, RYR2, and NXC1 mRNA expression, respectively, in nodal tissue. G) Plasma ELISA-measured concentration of epinephrine and norepinephrine. * P<0.0001 compared to control.

Hcn4:shJph2 mice demonstrate increased heart rates with blunted adrenergic modulation

To evaluate the impact of Jph2 knockdown on cardiac conduction and rhythm, Hcn4:shJph2 mice were first evaluated by surface ECG and ambulatory ECG telemetry. Hcn4:shJph2 mice exhibited a mildly increased sinus node recovery time (SNRT:123±3 ms) and QRS duration (8.7±0.3 ms) compared to controls (SNRT:111±1 ms, QRS:7.6±0.2 ms, respectively, P=0.02) consistent with intraventricular conduction delay. There was no change in PR, QT, and QTc intervals (Figure 2A and 2B, Sup Table I). Ambulatory ECG telemetry demonstrated no change in HR over a 24-hour period in Hcn4:shJph2 mice compared to control prior to tamoxifen-induction. However, following Jph2 knockdown, Hcn4:shJph2 mice demonstrated an increase in HR of 13.9% throughout the day from a mean of 583.6±14.4 bpm compared to controls of 512.1±13.6 bpm (P=0.003). This HR increase persisted throughout the sleep-wake cycle of the mice (Figure 2C and 2D). Taken together, these findings suggest increased SA nodal automaticity and intraventricular conduction delay without evidence of AV nodal block.

Figure 2:

ECG and telemetry analysis of Hcn4:shJph2 mice. A) Surface ECG and intracardiac EGM show atrial and ventricular depolarization in control and Hcn4:shJph2 mice. B) Bar graphs comparing the PR, QRS, and QT intervals. C and D) Plot and bar graph of telemetry monitored HR over a 24-hour period in control and Hcn4:shJph2 mice pre-tamoxifen (C, top) and post-tamoxifen (C, bottom). E) Bar graph of telemetry monitored change in HR with IP injection of beta-adrenergic agonists (epi, iso), antagonist (pro), muscarinic antagonists (atr) and agonist (car). *P<0.05 compared to control.

To test whether HR responsiveness to pharmacologic modulation was altered, telemetry-implanted mice were challenged with IP drug injections. Beta adrenergic stimulation (epinephrine, isoproterenol) caused increased HR in both Hcn4:shJph2 mice and controls. However, maximal HR achieved was blunted with Jph2 knockdown (618.6±10.5 bpm in Hcn4:shJph2 mice following epi, compared to 696.7±8.7 bpm in control, P=0.005; 741.1±0.5 bpm in Hcn4:shJph2 vs 770.8±8.9 bpm in control following iso, P=0.0002). Similarly, the HR-slowing effect of beta blockade (propranolol, 460.0±20.9 bpm in Hcn4:shJph2 vs 404.3±11.7 in control, P=0.02) and cardiac Na⁺/K⁺-ATPase blockade (digoxin, 413.7±15.4 bpm in Hcn4:shJph2 vs 320.5±32.6 bpm in control, P=0.03) was blunted compared to controls. This blunted modulation of HR was not seen with muscarinic receptor agonists (carbachol) and antagonists (atropine) (P=NS, respectively; Figure 2E). Hcn4:shJph2 mice demonstrated no spontaneous atrial ectopy or arrhythmias, ventricular ectopy or arrhythmias, or atrioventricular block on 24-hour telemetry. Taken together, these findings were consistent with increased automaticity of SA node with evidence for impaired responsiveness to beta-adrenergic modulation.

Hcn4:shJph2 mouse hearts demonstrate accelerated junctional rhythm in vivo and ex vivo

To evaluate intracardiac rhythm and arrhythmia inducibility, Hcn4:shJph2 mice were subjected to in vivo intracardiac EP studies. While control mice demonstrated sinus rhythm with clear electrical association between the atria and ventricles, knockdown mice demonstrated a high incidence of atrioventricular dissociation with ventricular depolarization occurring independently of the sinus node and faster than atrial depolarization without a change in QRS or ventricular depolarization morphology (Figure 3A). This was consistent with an accelerated junctional rhythm (AJR) driven by increased automaticity of the AV node. Spontaneous AJR was observed in 56% of Hcn4:shJph2 mice at baseline and was not observed in any of the control mice (P=0.004, Figure 3B). IP isoproterenol injection in knockdown mice triggered AJR in 78% of mice while no control mice demonstrated the arrhythmia (P<0.001, Figure 3C). Complete intracardiac EP study findings are summarized in the Sup Table I.

Figure 3:

Intracardiac EP and optical mapping of Hcn4:shJph2 mice. A) Surface ECG (I, II and III) and intracardiac EGMs (atrial, A; proximal ventricle, Vp; distal ventricle, Vd) recordings. Atrial EGM is expanded below with atrial depolarization (A) and ventricular depolarization (V) labeled. Scale bar, 250 ms. B and C) Bar graph of the proportion of spontaneous AJR (B) and adrenergically-induced AJR (C) on intracardiac EP monitoring. D) Ex vivo optical mapping demonstrating normal sinus rhythm in the control mice with depolarization (red) starting in the right atria (arrow) and moving from apex to base in the ventricle versus shJph2 mice which demonstrate ventricular depolarization without right atrial depolarization. Scale bar, 2 mm.

To confirm that this finding was intrinsic to the heart, and not secondary to non-cardiac influences such as sympathetic or parasympathetic innervation, hearts from Hcn4:shJph2 mice were Langendorff-perfused and ex vivo optical mapping was performed. Consistent with intracardiac EP studies, optical imaging of the heart’s electrical depolarization demonstrated spontaneous AJR manifest as ventricular depolarization without preceding right atrial depolarization (Figure 3D). This was confirmed by superimposition of myocardial voltage maps with visualization of the AV node at the crest of the intraventricular septum, which demonstrated AV node depolarization preceding ventricular depolarization without atrial depolarization (Sup Figure III). Taken together, these findings suggest that Hcn4:shJph2 mice have a predisposition to increased AV nodal automaticity which manifests as AJR.

Epicardial ex vivo rhythm monitoring of Hcn4:shJph2 hearts demonstrate junctional rhythm and inducible junctional tachycardia

To conclusively define the ex vivo cardiac rhythm, customized epicardial electrodes were placed directly on excised Hcn4:shJph2 heart atria and ventricles during retroaortic perfusion. Consistent with previous findings, 44% of Hcn4:shJph2 ex vivo hearts demonstrated AJR at baseline while none of the control hearts demonstrated AJR (P=0.04). Isoproterenol bolus into the cardiac perfusate induced a rapid onset junctional ectopic tachycardia (JET) with a rhythm significantly faster than AJR in 89% of knockdown mice while control mice demonstrated sinus rhythm (P<0.001, Figure 4A-F). To test whether antagonists of RyR2 opening terminated JET, we first treated mice with flecainide, which has been previously shown to reduce Ca²⁺ leak from RyR2 in cardiac myocytes without direct binding to RyR2¹⁹. Flecainide bolus into the perfusate did not terminate JET and frequently induced ventricular arrhythmias. In contrast, a similar perfusate bolus of a specific small molecule RyR2 inhibitor, EL20²⁰, abruptly terminated JET in 88% of knockdown mice with JET (P=0.01). In some cases, EL20 terminated successive episodes of isoproterenol-triggered JET in the same mouse with an overall conversion rate of 76% for all JET events (P=0.01, Figure 4G). These findings suggest that quickly administered beta agonists are able to trigger a rapid tachycardia from the AV node, consistent with JET, which can be converted to sinus rhythm with inhibition of RyR2.

Figure 4:

Epicardial EP monitoring and JET conversion with EL20 application. A and B) Ex vivo surface electrode EGM of control and Hcn4:shJph2 mice with atrial depolarization (A and vertical hash lines) and ventricular depolarization (V) labeled. C) Application of beta agonist into the cardiac perfusate (iso) induces rapid onset JET manifest as a narrow complex tachycardia with VA dissociation. D) Application of experimental RyR2 inhibitor EL20 terminates JET to normal sinus rhythm. E and F) Bar graphs demonstrating the proportion of knockdown mice with spontaneous AJR and inducible JET. G) Frequency of conversion of JET with flecainide (flec) and EL20 in each ex vivo heart (mice) and total JET events terminated (events). Vertical lines indicate atrial depolarizations. *P<0.05, **P<0.001 compared to control

Jph2 knockdown causes increased frequency and decreased amplitude of calcium transients in SA nodal calls

To determine whether intracellular Ca²⁺ signaling was associated with increased cell automaticity, we isolated SA nodal cells from Hcn4:WT and Hcn4:shJph2 mice and conducted single-cell Ca²⁺ imaging with Cal-520 fluorescent Ca²⁺ dye. Isolated SA nodal cells demonstrated classic spindle- or spider-like morphology (Figure 5A) and spontaneous depolarization activity manifest as Ca²⁺ transients. Hcn4:shJph2 SA nodal cells demonstrated a faster frequency of spontaneous transients; however, the transients were irregular and frequently prematurely terminated without traveling the length of the cell (Figure 5B). Ca²⁺ transient kinetics were also measured (Figure 5C). Overall, Hcn4:shJph2 SA nodal cells demonstrated an approximately 5-fold increase in transient frequency (P<0.0001, Figure 5D) with a reduce transient amplitude (P<0.0001, Figure 5E) and transient time-to-peak (P=0.02, Figure 5F). Finally, time constant of transient decay (tau) was found to be reduced in Hcn4:shJph2 SA nodal cells compared to controls (P=0.002, Figure 5G). Overall, these findings demonstrate that Jph2 knockdown is associated with increased spontaneous Ca²⁺ transient frequency that are reduced in amplitude and reduction in SR Ca²⁺ release.

Figure 5:

Jph2 knockdown causes increased nodal cell transient frequency while decreasing transient amplitude. A) Morphology of SA nodal cells in bright field and in LOI field. B) Representative confocal line scan images in Hcn4:WT (left, top) and Hcn4:shJph2 (right, top), and corresponding Ca²⁺ fluorescent tracings (bottom) of observed Ca²⁺ transients at noted positions in line scan images (a, b, and c). C) Representative fluorescence tracings for individual Ca²⁺ transients in Hcn4:WT (left) and Hcn4:shJph2 nodal cells. D-G) Histograms demonstrating the frequency of spontaneous transients (D), transient amplitude (E), transient time-to-peak (F), and time constant of transient decay (tau) (G) in Hcn4:WT mice (white fill) vs. Hcn4:shJph2 mice (black fill). Cell number per analysis is noted in the respective bars. Cells which were derived from 7 Hcn4:WT (Ctrl) and 4 Hcn4:shJph2 mice, respectively. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 compared to control.

Jph2 knockdown causes increased store-calcium leak

In order to determine whether this increase in Ca²⁺ transient frequency was associated with increased Ca²⁺ leak from the SR Ca²⁺ store, we measured caffeine-triggered SR Ca²⁺ release and measured Ca²⁺ sparks. Hcn4:shJph2 SA nodal cells demonstrated a markedly reduced level of SR Ca²⁺ release compared to the control cells (Figure 6A-C). Calcium sparks were more frequent (P<0.0001) in Hcn4:shJph2 cells; however, demonstrated a smaller amplitude (P=0.01), a trend toward smaller width (P=0.06), and duration (P=0.008) compared to controls (Figure 6D-H). As the result, the Jph2 knock down cells were calculated to have an approximately 20% higher amount of Ca²⁺ leak from SR compared to the control cells (P=0.001, Figure 6I). Taken together, the results suggest that increased Ca²⁺ leak from the SR leads to reduced levels of stored Ca²⁺ and is driven by increased RyR2 opening.

Figure 6:

Jph2 knockdown causes decreased ST-stored Ca²⁺ with increased Ca²⁺ spark frequency and SR Ca²⁺ leak. A-B) Representative line scan images (top) and corresponding Ca²⁺ fluorescent tracings (bottom) of Cal-520 loaded Hcn4:WT (A) and Hcn4:shJph2 (B) nodal cells exposed to caffeine. C) Histogram of store Ca²⁺ release following caffeine exposure. D) Representative line scan images of Ca²⁺ sparks in Hcn4:WT (left) vs. Hcn4:shJph2 (right) mice. E-I) Histograms demonstrating Ca²⁺ spark amplitude (E), width (full width at half maximum amplitude, FWHM) (F), duration (full duration at half-maximum, FDHM (G), CaSpF frequency (H), and calculated Ca²⁺ leak (I) in Hcn4:WT mice (white) vs. Hcn4:shJph2 (black fill) calculated based on the equation FWHM×FDHM×F/F₀×CaSpF. Cell number per analysis is noted in the respective bars. Cells which were derived from 7 Hcn4:WT (Ctrl) and 4 Hcn4:shJph2 mice, respectively. *P<0.05 *P<0.05, **P<0.01, ****P<0.0001.

Jph2 knockdown is associated with increased rate of diastolic depolarization and increased cellular automaticity in SA nodal cells

To determine whether the increased transient frequency and store Ca²⁺ leak was associated with increased SA nodal cell automaticity, we recorded spontaneous action potentials (AP) in isolated single SA nodal cells. Compared to Hcn4:WT nodal cells, the firing rate of AP was significantly higher in Hcn4:shJph2 nodal cells (355±13 vs. 256±22 bpm, P=0.003, respectively) and the AP cycle length was markedly shorter (171±6 vs. 244±22 ms, P=0.005). The rate of the diastolic depolarization (DDR) also significantly higher in shJPH2 nodal cells (224±21 vs. 159±9 mV/s, P=0.04) and the duration of the diastolic depolarization (DI) was shorter in shJph2 cells (101±7 vs. 152±14, P=0.007, Figure 7A-F). Conversely, there were no significant changes in the maximum diastolic potential (MDP), the AP threshold potential (E_-th), the AP threshold (V_-th), AP amplitude (APA), AP duration, and dV/dt_max (Sup Figure IV). Moreover, some of the spontaneous membrane depolarizations were terminated early and failed to burst into action potentials (Sup Figure V), which was consistent with the observation that some Ca²⁺ transients were prematurely terminated without traveling the length of the nodal cell. Overall, these findings suggested increased automaticity in shJph2 SA nodal cells.

Figure 7:

Jph2 knockdown causes increased cellular automaticity and reduced sodium-calcium exchanger current (I_NCX) in SA nodal cells. A and B) Representative action potential recordings of SA nodal cells in Hcn4:shJph2 knock-down mice and Hcn4:WT control mice. C-F) Graphs of the action potential parameters including the AP firing rate (C), diastolic depolarization (DDR, panel D), action potential cycle length (CL, E), and the duration of diastolic depolarization (DI, F). G) Representative traces of I_NCX in Hcn4:WT (blue) vs. Hcn4:shJph2 (red) primary SA nodal cells. H and I) Graph of the mean values of I_NCX currents of SA nodal cells measured at −60mV (H) and +80mV (I) in Hcn4:WT (blue) vs. Hcn4:shJph2 (red) SA nodal cells. Each cell is identified by an independent data point and were derived from >4 mice. *P<0.05 compared to control. **P<0.01 compared to control. NS, no significant difference.

Jph2 knockdown causes reduced I_NCX

Given the increase in Ca²⁺ leak observed, and the established interplay between intracellular Ca²⁺ leak and Ncx function in ventricular myocytes²¹, we next performed patch clamp analysis of I_NCX. We found that that I_NCX was reduced by approximately 50% in Hcn4:shJph2 nodal cells compared to control at −60 mV (P=0.03). Further, there was no significant difference in I_NCX at +80 mV (Figure 7G-I). Overall, this suggests that loss of Jph2 expression in nodal cells is associated with decreased Ncx function during resting membrane potentials.

JPH2 gene locus may be associated with clinical cases of nodal disease

To explore the possibility that the JPH2 locus may be associated with nodal disease in humans, we next pursued PheWAS analysis of coding variants in JPH2 using the BioVU platform. We identified 18 nominally significant associations with a P-value <0.01, including identified associations between variant rs3810510 (p.Ala396Thr) and several clinical diagnoses of SA and AV nodal disease. Among the top 13 phecodes were four cognate human diagnoses of phenotypes observed in our mouse experiments, including first degree AV block (OR 1.49, P=0.003), AV block (OR 1.26, P=0.004), SA node dysfunction (OR1.23, P=0.006) and Mobitz II AV block (OR 1.98, P=0.006). These results are summarized in Table 1 with a full list of the top PheWAS results summarized in Sup Table II. These PheWAS findings suggest an important role JPH2 in nodal disease in humans, although further studies are needed to establish this association.

Table 1:

Top phecodes associated with coding exons in JPH2

Phecode	SNP	OR	P-value	N cases	N controls
First degree AV block	rs3810510.42747247.C.T_T	1.50	0.003	228	3775
Atrioventricular block	rs3810510.42747247.C.T_T	1.26	0.004	937	3775
Sinoatrial node dysfunction (bradycardia)	rs3810510.42747247.C.T_T	1.23	0.006	1183	3775
Mobitz II AV block	rs3810510.42747247.C.T_T	2.00	0.006	50	3775

SNP, single nucleotide polymorphism; OR, odds ratio; AV, atrioventricular

Discussion

Store-calcium leak is associated with increased nodal automaticity

Action potential generation of the SA node is an automatic, yet highly regulated, event that is orchestrated by a complex interplay between the membrane voltage and Ca²⁺ clocks. Recent evidence from multiple investigators suggests that there is an intracellular Ca²⁺ clock which plays a critical role in impulse generation through rhythmic, spontaneous release of Ca²⁺ from intracellular SR stores via RyR2⁶. Here, we find that reduced Jph2 expression is associated with faster spontaneous nodal cell depolarizations and increased SR store Ca²⁺ leak in vitro, which manifests as sinus tachycardia and AV nodal tachycardia, in vivo. The identification of reduced I_NCX at negative membrane potentials in Hcn4:shJph2 mice, which is similar to controls a positive membrane potentials, suggests reduced Ca²⁺ efflux from the cell via Ncx during resting membrane states. This fits with a possible “Ca²⁺ overload” state that is driving increased cellular automaticity. In support of this, application of ryanodine to nodal cells, an inhibitor of RyR2 opening, causes slower heart rate²². In addition, build-up of SR store Ca²⁺ has been linked to recruitment of depolarizing currents which are a part of the membrane voltage clock, such as I_f created by Hcn4, I_L-Ca from the L-type Ca²⁺ channel (Ca_V1.2), and I_T-Ca from the T-type Ca²⁺ channels (Ca_V3.1 and 3.2). These currents serve to increase membrane potential spontaneously and coordinate to drive nodal depolarization²³. In this way, it is believed that perturbed Ca²⁺-signaling within the nodal cells, such as increased SR Ca²⁺ leak from RyR2, may result in loss of proper automaticity and nodal disease. We see evidence for this in both nodal cell Ca²⁺ transients and spontaneous APs which occur more rapidly yet can terminate before reaching either the full length of the cell or fully depolarize the cell, respectively. It is possible that this process can override the typical responsiveness of the nodal tissue to adrenergic and cholinergic regulation. Beta adrenergic stimulation of protein kinase A has been linked to increased nodal depolarization secondary to increased store Ca²⁺ release²⁴. While we find that the heart rate responsiveness of Hcn4:shJph2 mice to chronotropic drugs, primarily adrenergic receptor agonists and antagonists, is blunted, additional studies are needed to confirm whether there is a difference in the impact of cholinergic versus adrenergic drugs on intracellular Ca²⁺ handling.

Jph2 negatively regulates RyR2-mediated calcium leak

While our study is the first to explore the role of Jph2 in nodal cells Jph2 has a well-established role in regulating Ca²⁺ release from the SR in ventricular myocardial cells, among other critical roles in the heart. Jph2 is a muscle-specific protein located in the junctional membrane complex (JMC) of the myocyte which plays a critical role in maintaining efficient intracellular Ca²⁺-signaling for myocyte contraction as well as proper gating of RyR2²⁵. As part of the JMC, Jph2 colocalizes with several sarcolemmal proteins and SR proteins that create the critical cellular ultrastructure needed for efficient Ca²⁺-signaling²⁶. Jph2 directly interacts with RyR2, and expression silencing results in increased RyR2 gating by lipid bilayer analysis and increased Ca²⁺ leak from the SR manifest as increased Ca²⁺ spark size and frequency^{12, 21}. Further, in the myocyte Jph2 also interacts with Speg, which can indirectly alter RyR2 and the SR Ca²⁺-ATPase by phosphorylation and activity^27-29. While this aberrant Ca²⁺ homeostasis has been linked with the development of cardiomyopathic disease and heart failure, it also serves as an arrhythmic substrate and can generate cellular ectopy and spontaneously depolarizing myocardial cells^10-12. In this study, we find that Jph2 expression knockdown in the nodal cells causes reduced Ca²⁺ transient amplitude while increasing automaticity of the cell. This is similar to in vitro and in vivo models of cardiac myocyte models of Jph2 expression silencing^10-12. Further, our observation that Hcn4:shJph2 isolated nodal cells demonstrate leak of SR store Ca²⁺, which is driven by increased spark frequency, supports an analogous role for Jph2 in the negative gating regulation of RyR2 seen in cardiac myocytes. Our finding that Ncx1 function is impaired during “resting” membrane potentials is in line with previous work identifying a similar reduction in Jph2 knockdown within cardiac myocytes²¹. This impairment may reflect a regulatory effect of Jph2 on Ncx1. This possibility is supported by evidence that loss of Ncx1 expression in murine nodal cells results in rapid nodal depolarizations and an increase in automaticity³⁰. Further, Ca²⁺ leak-driven automaticity is supported by resolution of tachycardic arrhythmias by EL20, a direct RyR2 antagonist. EL20 is a novel tetracaine-derivative that has been shown to diminish arrhythmia burden in a murine model of catecholaminergic polymorphic ventricular tachycardia²⁰. While its role on nodal tissue is unexplored, it is believed to reduce the diastolic leak of SR-stored Ca²⁺ by selectively inhibiting RyR2 channels in cardiac myocytes²⁰. Given the emerging role of abnormal Ca²⁺ homeostasis and signaling in driving automaticity, our findings suggest that targeting SR-mediated Ca²⁺ leak may be an avenue for the treatment of nodal arrhythmias with increased automaticity.

Calcium clock regulation in nodal cells

The role of SR-mediated Ca²⁺ leak in nodal cell automaticity remains complex. Mutations in RYR2, which typically cause increased opening probability of the channel, and resultant leak of SR Ca²⁺, are a major cause of catecholaminergic polymorphic ventricular tachycardia (CPVT). While the majority of mouse models hosting RYR2 mutations associated with arrhythmias focus on inducible ventricular arrhythmias in their characterization, there is a small body of work exploring the impact of these mutations on nodal cell automaticity. Interestingly, these mutations are associated with sinus bradycardia and reduced nodal cell automaticity. In one isolated SA nodal cells from the RYR2-R4496C mouse demonstrated reduced Ca²⁺ transient frequency in the setting of increased SR Ca²⁺leak³¹. The authors also identified reduced L-type Ca²⁺ current leak and reduced SR store Ca²⁺ levels and conclude that it is this reduction in overall Ca²⁺ load the reduces the Ca²⁺ clock control of automaticity. This phenomenon is also present in patients with CPVT for whom RYR2 mutations result in RYR2-mediated Ca²⁺ leak from the sarcoplasmic reticulum. Rather than sinus tachycardia, which would be expected given the findings from these mice, humans consistently demonstrated sinus bradycardia associated with RYR2 pathologic mutations³². This divergence in phenotype between mouse and human nodal automaticity is also supported by our PheWAS analysis of JPH2, which identified an association between this locus and bradycardic sinus node dysfunction in patients. In at least one genetic model of CPVT, both patients and a knock-in mouse hosting the RYR2-R420Q variant demonstrated bradycardia through impaired Ca²⁺ clock signaling and what the authors suggest is impaired beta adrenergic modulation of heart rate³³. This raises the possibility that extracardiac modulators of heart rate, such as increased sympathetic regulation in patients CPVT or in patients with significant bradycardia, may blunt the effect of the intracellular Ca²⁺ clock. This possibility remains largely untested. Finally, while there are clear species-specific differences in nodal cell automaticity response due to inhibition of store Ca²⁺ release, correlative studies to explore variability in humans are largely lacking³⁴. Thus, the divergence of these phenotypes remains an open question.

While canonically viewed as a structural protein which plays a role in excitation-contraction coupling and intracellular Ca²⁺ regulation, it has been recently shown that Jph2 plays a critical role in driving the ultrastructural cellular remodeling of cardiac myocytes. Stress-induced calpain-mediated cleavage of Jph2 creates an N-terminal fragment which translocate to the nucleus and initiates a cellular response to this stress³⁵. While the transcriptional role of Jph2 in nodal tissue of the heart remains unknown, it is plausible Jph2 may play a role beyond purely direct molecular coupling. It is also possible that other known regulators of the Ca²⁺ clock play still-undefined roles in the observed increased in automaticity; however, future studies are needed to determine this.

Limitations

This study has several limitations. Due to inability to consistently detect atrial depolarizations, we were unable to rigorously determine whether junctional rhythm, or JET, occurred in ambulatory mice using telemeters. It remains unclear whether EL20 blocks any non-RyR2 targets that may play a role in reducing the nodal automaticity observed with Jph2 knockdown. It also remains unclear what the role of the membrane clock is in modulating the automaticity and Ca²⁺ transients observed. Experimentally, measurement of SR Ca²⁺ store in a spontaneously beating nodal cell may be affected by the presence and timing of Ca²⁺ transients at the time of caffeine exposure. The role that NCX current, as well as the kinases PKA and CaMKII, may have in modulating nodal automaticity in this model remains an area of future research. Finally, due to the large number phecodes comprising the PheWAS study, adjustment for multiple comparisons was not performed. Thus, the association between variants in JPH2 and the diseases noted should be considered hypothesis-generating and require further validation to establish a true relationship.

Conclusions

Nodal-specific knockdown of Jph2 expression causes increased nodal automaticity through increased Ca²⁺ leak from intracellular stores. This store Ca²⁺ leak drives faster and more irregular spontaneous nodal depolarizations which increased nodal automaticity and manifest as tachycardia and junctional rhythms in mice. Dysregulated intracellular Ca²⁺ underlies nodal arrhythmogenesis in this mouse model.

What is Known:

• An intracellular “calcium clock” can regulate the spontaneous automaticity of cardiac nodal cells.

What the Study Adds:

• Reduced Jph2 expression in murine cardiac nodal cells results in increased nodal automaticity and increased leak of calcium from the sarcomplasmic reticulum.

• In vivo expression silencing of Jph2 in mice results in sinus tachycardia and accelerated junctional rhythm, reflecting increased nodal automaticity.

• The JPH2 gene locus is associated with nodal disease in humans by phenome-wide association.

Nonstandard Abbreviations and Acronyms

AJR

accelerated junctional rhythm

AP

action potential

APA

action potential amplitude

AV

atrioventricular node

CaMKII

calmodulin-dependent protein kinase II

DDR

diastolic depolarization rate

DI

duration of diastolic depolarization

ECG

electrocardiogram

E_-th

action potential threshold potential

EP

electrophysiology

FWHM

full width at half maximum amplitude

FDHM

full duration at half maximum amplitude

GFP

green fluorescent protein

Hcn4

Hcn4-driven cre-recombinase

HR

heart rate

IJ

internal jugular

JET

junctional ectopic tachycardia

JMC

junctional membrane complex

Jph2

junctophilin 2

mTmG

membrane-targeted tdTomato and green fluorescent protein

MDP

maximum diastolic potential

NCX/Ncx1

sodium-calcium exchanger

PKA

protein kinase A

PheWAS

phenome-wide association study

RFP

red fluorescent protein

RyR2

ryanodine receptor type 2

SA

sinoatrial node

ShJph2

Jph2-targeted short-hairpin

SNRT

sinus node recovery time

SR

sarcoplasmic reticulum

V_-th

action potential threshold