Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Feb 18.
Published in final edited form as: J Am Coll Cardiol. 2013 Nov 27;63(6):549–559. doi: 10.1016/j.jacc.2013.10.062

Coxsackie and adenovirus receptor (CAR) is a modifier of cardiac conduction and arrhythmia vulnerability in the setting of myocardial ischemia

Roos FJ Marsman 1, Connie R Bezzina 1, Fabian Freiberg 2, Arie O Verkerk 3, Michiel E Adriaens 1, Svitlana Podliesna 1, Chen Chen 2, Bettina Purfürst 4, Bastian Spallek 5, Tamara T Koopmann 1, Istvan Baczko 6, Cristobal G dos Remedios 7, Alfred L George Jr 8, Nanette H Bishopric 9, Elisabeth M Lodder 1, Jacques MT de Bakker 1,10, Robert Fischer 5, Ruben Coronel 1, Arthur AM Wilde 1, Michael Gotthardt 2,11, Carol Ann Remme 1
PMCID: PMC3926969  NIHMSID: NIHMS542095  PMID: 24291282

Abstract

Objectives

To investigate the modulatory effect of the Coxsackie and adenovirus receptor (CAR) on ventricular conduction and arrhythmia vulnerability in the setting of myocardial ischemia.

Background

A heritable component in risk for ventricular fibrillation (VF) during myocardial infarction (MI) has been well established. A recent genome-wide association study (GWAS) for VF during acute MI has led to the identification of a locus on chromosome 21q21 (rs2824292) in the vicinity of the CXADR gene. CXADR encodes the coxsackie and adenovirus receptor (CAR), a cell adhesion molecule predominantly located at intercalated discs of the cardiomyocyte.

Methods

The correlation between CAR transcript levels and rs2824292 genotype was investigated in human left ventricular samples. Electrophysiological studies and molecular analyses were performed CAR haploinsufficient mice (CAR+/−).

Results

In human left ventricular samples, the risk allele at the chr21q21 GWAS locus was associated with lower CXADR mRNA levels, suggesting that decreased cardiac levels of CAR predispose to ischemia-induced VF. Hearts from CAR+/− mice displayed ventricular conduction slowing in addition to an earlier onset of ventricular arrhythmias during the early phase of acute myocardial ischemia following LAD ligation. Connexin43 expression and distribution was unaffected, but CAR+/− hearts displayed increased arrhythmia susceptibility upon pharmacological electrical uncoupling. Patch-clamp analysis of isolated CAR+/− myocytes showed reduced sodium current magnitude specifically at the intercalated disc. Moreover, CAR co-precipitated with NaV1.5 in vitro, suggesting that CAR affects sodium channel function through a physical interaction with NaV1.5.

Conclusion

We identify CAR as a novel modifier of ventricular conduction and arrhythmia vulnerability in the setting of myocardial ischemia. Genetic determinants of arrhythmia susceptibility (such as CAR) may constitute future targets for risk stratification of potentially lethal ventricular arrhythmias in patients with coronary artery disease

Keywords: arrhythmia, ventricular fibrillation, ischemia, single nucleotide polymorphism genetics, ion channels

INTRODUCTION

Ventricular fibrillation (VF) is a frequent and potentially lethal complication of acute myocardial infarction (MI). In this setting, VF is the consequence of disturbed electrical properties of the ischemic myocardium, which includes a decrease in cardiomyocyte excitability and cell-to-cell coupling. These factors result in conduction slowing of the cardiac electrical impulse, which is a prerequisite for the occurrence of life-threatening arrhythmias such as VF in MI (1).

Traditional cardiovascular risk factors do not identify which MI patient is at risk for VF, and specific and sensitive risk predictors are currently lacking. A heritable component in the determination of risk for VF and sudden cardiac death during MI has been well established (2, 3), but progress in understanding the molecular and genetic determinants of ischemia-induced VF has been limited. In order to identify susceptibility genes, our group recently conducted a genome-wide association study, which led to the identification of common genetic variants at the chromosome 21q21 locus associated with risk of VF in the setting of acute MI (4). The most significant association signal for VF on the chromosome 21q21 locus comprised the single nucleotide polymorphism (SNP) rs2824292 located immediately upstream of two genes, CXADR and BTG3 (the only genes within a region of one megabase spanning the association signal) (4). The BTG3 gene encodes B cell translocation gene 3, a member of the anti-proliferative BTG/Tob protein family, known to regulate cell cycle progression, gene expression, tumorigenesis and cancer (5). The CXADR gene encodes the Coxsackie and adenovirus receptor (CAR), a transmembrane cell adhesion molecule predominantly located at the intercalated disc between cardiomyocytes (6–9). While CAR has been recognized primarily for its involvement in virus-mediated myocarditis, recent studies in CAR knockout mice revealed an atrio-ventricular conduction slowing due to loss of expression of the gap junction protein connexin-45 (Cx45) (9, 10). Moreover, expression levels of Cx43, the predominant gap junction molecule in ventricular myocardium, were decreased in hearts from CAR knockout mice (9), raising the possibility that CAR may play a role in ventricular conduction. Of note, CAR is known to be differentially regulated in various cardiac disease states, including dilated cardiomyopathy and MI (6, 11).

We hypothesized that CAR impacts on ventricular conduction and susceptibility to ventricular arrhythmia in the setting of MI. We demonstrate that the rs2824292 risk genotype is associated with decreased CXADR expression in human heart. Furthermore, we show that mice haploinsufficient for CAR display ventricular conduction slowing and an earlier onset of ventricular arrhythmias during MI, at least in part mediated by a reduced sodium current magnitude at the intercalated disc.

METHODS

The methods section can be found in the Online data supplement.

RESULTS

Rs2824292 regulates cardiac expression of CXADR, the gene encoding CAR

Since the risk haplotype tagged by rs2824292 occurs in a non-coding region of chr21q21, it is likely to impact on VF risk through effects on gene expression. We therefore investigated whether rs2824292 genotype correlates to transcript expression levels of the only two neighboring genes within one megabase distance, namely CXADR (located 100 kb downstream of rs2824292) and BTG3 (located 179 kb downstream). Assessment of transcript abundance for these genes in normal donor-heart myocardium (n=129) uncovered an association between genotype at rs2824292 and cardiac CXADR mRNA levels (Figure 1). Individuals carrying one or two copies of the risk allele (AG or GG genotype) displayed significantly lower (0.6 fold) CXADR mRNA expression compared to individuals with the non-risk (AA) genotype (dominant model p=0.002; additive model p=0.006). No such relation was observed with cardiac expression of BTG3. Thus, the VF risk allele at rs2824292 is associated with decreased cardiac expression of CXADR, suggesting that decreased levels of CAR predispose to VF during ischemia.

Figure 1. Rs2824292 regulates cardiac CXADR levels.

Figure 1

Open in a new tab

(A) The human 21q21 locus. CXADR and BTG3 are located within 500 kb region downstream of rs2824292. (B) Mean mRNA expression of CXADR and BTG3 as a function of genotype at rs2824292 in human heart samples, standardized to the geomean of cTNI and HPRT mRNA expression levels (* p=0.05; ** p< 0.01). Mean expression levels in hearts from AG heterozygotes and GG homozygotes are normalized to the mean expression levels from hearts of AA homozygotes.

CAR haploinsufficient mice (CAR+/−)

To investigate the effects of reduced CAR expression on cardiac conduction and arrhythmogenesis, we generated mice deficient for CAR (Supplemental Figure 1 and ref (12)) Homozygous CAR deficient mice were not viable but died in mid-gestation between day E10.5 and E12.5 (Supplemental Figure 2 and Supplemental Table 1). Heterozygous CAR deficient mice (CAR+/− ) were born at the expected Mendelian ratio (Supplemental Table 1) and had a normal life-span.

Earlier onset of inducible ventricular arrhythmias during acute myocardial ischemia in CAR+/− mouse hearts

To investigate whether reduced CAR levels predispose to arrhythmia during acute myocardial ischemia, we induced regional ischemia in Langendorff-perfused wild-type (WT) and CAR+/− hearts by ligating the left anterior descending (LAD) artery and tested arrhythmia inducibility. Ischemic zone size was not significantly different between WT and CAR+/− (Figure 2A). Hearts were stimulated from the non-ischemic area at a basic cycle length (BCL) of 120 ms (S1), and arrhythmia inducibility was tested before LAD ligation and at different intervals during ischemia using up to three extrastimuli (S2-S3-S4) and burst pacing. Spontaneous sustained arrhythmias did not occur in WT or CAR+/− hearts at baseline or during the ischemic period. Arrhythmias were never induced prior to LAD ligation, but were induced in all WT and CAR+/− hearts during ischemia. These included premature ventricular contractions (PVC), non-sustained ventricular tachycardia (VT) and sustained VT (Figure 2B). During the first 15 minutes of ischemia, we observed inducible arrhythmias (non-sustained and sustained) in 7 out of 9 CAR+/− hearts. In contrast, arrhythmias were induced in only 1 out of 8 WT hearts (p<0.05) (Figure 2C and D). Similarly, in the first 15 minutes of ischemia there was a trend towards a higher prevalence of inducible sustained VTs (> 3 seconds) in CAR+/− compared to WT hearts (Figure 2E). During later stages of the ischemic episode, arrhythmia inducibility was not different between WT and CAR+/− hearts. Thus, CAR+/− hearts displayed an advanced onset of ventricular arrhythmia inducibility during the early phase of acute myocardial ischemia.

Figure 2. Accelerated onset of inducible arrhythmias during myocardial ischemia in isolated CAR+/− hearts.

Figure 2

Open in a new tab

(A) Ischemic zone size after LAD ligation (non-perfused area; as a percentage of the left ventricle) was similar in WT and CAR+/− hearts. (B) Arrhythmia inducibility during LAD ligation. Typical examples of ventricular electrograms show extrastimuli (S1-S4) followed by either no arrhythmia (upper panel), six ventricular extrasystoles (premature ventricular contractions [PVC] - middle panel), or sustained ventricular tachycardia (lower panel). (C) Kaplan Meier plot showing an accelerated onset of first arrhythmic event during myocardial ischemia in CAR+/− hearts (n=9) as compared to WT (n=8). (D,E) CAR+/− hearts display increased inducibility of all types of arrhythmia (D) and of sustained ventricular tachycardia (VT > 3 seconds) (E) during the first 15 minutes of ischemia (*p<0.05).

Ventricular conduction slowing in isolated CAR+/− hearts

We hypothesized that altered electrophysiological properties secondary to CAR deficiency renders CAR+/− hearts more susceptible to arrhythmia inducibility in the setting of myocardial ischemia. Surface ECG analysis demonstrated no significant differences in ECG parameters (RR-, PQ-, QRS-, and QTc-intervals) between CAR+/− and WT mice (Supplemental Figure 3). We therefore studied ventricular conduction in more detail by electrical mapping of isolated WT and CAR+/− hearts. A 247-point electrode was placed on the left ventricular (LV) epicardial surface and hearts were stimulated from the center of the electrode. Local electrograms were used to construct activation maps (Figure 3A), from which longitudinal (CVL) and transversal (CVT) conduction velocities were determined. Stimulation threshold and shortest coupled intervals of S2, S3 and S4 (effective refractory period, ERP) were not significantly different between CAR+/− and WT hearts (Supplemental Table 3). During basic stimulation (S1), CVL (CAR+/− : 63.7±4.7; WT 77.8±2.8; p=0.02) and CVT (CAR+/−: 28.6±2.0; WT 38.0±2.5; p=0.01) were significantly reduced in CAR+/− compared with WT (Figure 3B-C). These differences in conduction velocity were also observed after application of three extrastimuli (S4; CAR+/− : 42.5±4.6; WT 55.5±4.2; p=0.054) and CVT (CAR+/− : 28.6±1.8; WT 33.9±1.5; p=0.04) (Figure 3B-C).

Figure 3. Ventricular conduction slowing in CAR+/− hearts.

Figure 3

Open in a new tab

(A) Representative left ventricular epicardial activation maps at baseline and after carbenoxolone (CBX) infusion in WT and CAR+/− hearts. Crowding of isochrones is indicative of conduction slowing. (B,C) Average longitudinal (B) and transversal (C) conduction velocities measured during basic stimulation (120 ms) and three extrastimuli (S4) at baseline and after CBX (* p<0.05 CAR+/− vs. WT; # p<0.05 CBX vs. baseline).

Pharmacological electrical uncoupling induces ventricular arrhythmias in CAR+/− hearts

Conduction slowing in CAR+/− hearts may stem from altered tissue architecture, cell-to-cell coupling, and/or cardiomyocyte excitability. No evidence for myocardial hypertrophy or fibrosis was found in CAR+/− hearts (Supplemental Figure 4). In addition, transmission electron microscopy did not reveal ultrastructural changes in CAR+/− hearts (Supplemental Figure 5). To evaluate whether the conduction slowing in CAR+/− hearts is mediated by alterations in gap junction composition and/or function, we investigated the expression and distribution of connexins and challenged WT and CAR+/− hearts with the gap junction uncoupler carbenoxolone. Carbenoxolone is known to decrease conduction velocity both in the longitudinal and transversal direction without affecting action potential characteristics and underlying ionic currents (13). During stimulation at basic cycle length (S1), carbenoxolone significantly reduced CVL in both WT and CAR+/− hearts (Figure 3B). Similarly, CVT was decreased by carbenoxolone in WT and CAR+/− hearts after application of 3 extrastimuli (S4; Figure 3C).

The consequences of the observed conduction slowing after carbenoxolone infusion with respect to arrhythmia inducibility were tested using up to three extrastimuli (S2-S3-S4) and burst pacing. The shortest possible coupling intervals of S2-S3-S4 and burst trains were not significantly different between CAR+/− and WT hearts at baseline and after carbenoxolone infusion (Supplemental Table 3). Spontaneous arrhythmias did not occur in either experimental group and we were unable to induce arrhythmias in WT or CAR+/− hearts at baseline. However, while none of the WT hearts showed induced arrhythmias upon carbenoxolone infusion, 50% of CAR+/− hearts developed sustained ventricular arrhythmias during burst pacing following carbenoxolone administration (CAR+/− 4/8 vs. WT 0/9, p=0.015) (Figure 4A-B). VTs induced in CAR+/− hearts following carbenoxolone infusion were reproducible and lasted for more than 5 seconds to 1 minute. Activation maps of two consecutive beats of an induced ventricular tachycardia in a CAR+/− heart show the impulse propagating around a line of block, indicating a re-entrant mechanism underlying the arrhythmia (Figure 4C). Ventricular mRNA and protein levels of the gap junction proteins Cx43 and Cx45 were not different between WT and CAR+/− hearts (Figure 5A-C) and there were no differences in Cx43 distribution (in particular, no lateralization or redistribution) in ventricular cryosections (Figure 5D) and isolated myocytes (Supplemental Figure 6). Protein expression levels of other intercalated disc proteins, namely β-catenin, zonula occludens-1 (ZO-1), N-cadherin and Cx45, were not different between WT and CAR+/− hearts (Supplemental Figure 7).

Figure 4. Increased arrhythmia inducibility after carbenoxolone infusion in CAR+/− hearts.

Figure 4

Open in a new tab

(A) Example of a sustained ventricular arrhythmia induced by burst pacing in a CAR+/− heart after infusion of carbenoxolone (CBX). (B) Increased incidence of inducible arrhythmias after CBX in CAR+/− hearts as compared to WT (p=0.015). (C) Activation maps of two consecutive VT beats induced in a CAR+/− heart showing the impulse propagating around a line of functional conduction block, indicating a re-entrant mechanism. The right panel depicts electrograms at positions A-B-C-A’ which correspond to recording sites as indicated in the activation maps (left). Local activation times are determined as the maximal negative dV/dt in each electrogram (indicated by red dots). BCL: Basic cycle length (120 ms).

Figure 5. Unaltered Cx43 and NaV1.5 expression and distribution in CAR+/− hearts.

Figure 5

Open in a new tab

(A) Left ventricular Cx43, Scn5a and Car mRNA levels (n=4 per group, *p<0.05). (B) Immunoblots of NaV1.5, Cx43 and CAR (whole cell and membrane fraction; calnexin as a loading control). (C) NaV1.5, Cx43 and CAR proteins levels (normalized to WT littermate control; n=5 mice in each group, *p<0.05) (D) Immunofluorescent stainings of Cx43 and NaV1.5 in left ventricular tissue; pan-cadherin (Pan-cadh) and N-cadherin (N-cadh) used as intercalated disc markers.

Reduced sodium channel availability in CAR+/− mice and cardiomyocytes

Since the cardiac sodium channel protein NaV1.5 is enriched in the intercalated disc region, CAR may also impact on conduction by affecting sodium channel expression and/or function. Scn5a mRNA and NaV1.5 protein levels were not different between CAR+/− and WT ventricular tissue (Figure 5A-C), and there were no differences in NaV1.5 localization at the intercalated disc or lateral membrane (Figure 5D). Functionally however, action potential (AP) upstroke velocity (dV/dtmax, a measure of sodium channel availability under nearphysiological conditions (14)) was altered in cardiomyocytes isolated from CAR+/− hearts (Figure 6A,B). On average, dV/dtmax was ≈15% lower in CAR+/− myocytes as compared to WT, indicating a decrease in functional sodium channel availability secondary to CAR haploinsufficiency. Resting membrane potential, AP amplitude and AP duration were unchanged in CAR+/− cardiomyocytes (Figure 6B). In vivo acute administration of the sodium channel blocker flecainide in anaesthetized mice revealed a significantly larger increase in PR-interval and QRS-duration in CAR+/− versus WT mice, underlining the functional relevance of the reduced sodium channel availability for cardiac conduction (Supplemental Figure 3C,D).

Figure 6. Reduced sodium channel availability and decreased sodium current amplitude at the intercalated disc in CAR+/− cardiomyocytes.

Figure 6

Open in a new tab

(A) Action potentials (APs) measured from isolated left ventricular cardiomyocytes of WT and CAR+/− hearts (inset: dV/dt of the AP upstroke). (B) Average AP characteristics. Maximal AP upstroke velocity (dV/dtmax) was significantly lower in CAR+/− myocytes as compared to WT (* p<0.05). Resting membrane potential (RMP); action potential amplitude (APA) and action potential duration at 20, 50 and 90% of repolarization (APD20, APD50 and APD90, respectively) were not different between WT and CAR+/−. (C,D) Current-voltage (I-V) relationships show similar INa amplitude at the lateral membrane of WT and CAR+/− cardiomyocytes (C), but reduced INa amplitude in CAR+/− versus WT at the intercalated disc (assesed by two-way RM ANOVA followed by Holm-Sidak post-hoc testing: p=0.002 for overall effect) (D). Asterisks (*)denote membrane potentials at which INa amplitude is significantly different between WT and CAR+/−. (E,F) Average voltage-dependencies of activation and inactivation demonstrate no significant difference between WT and CAR+/− at the lateral membrane (E) or at the intercalated disc (ID) (F).

CAR haploinsufficiency affects sodium current magnitude specifically at the intercalated disc

Sodium channels are located both at the lateral membrane and at the intercalated disc of the cardiomyocyte (15, 16). Since CAR is localized predominantly at the intercalated disc, we hypothesized that CAR deficiency affects sodium current (INa) preferentially in this region. We therefore investigated INa characteristics at the lateral membrane and at the intercalated disc of isolated CAR+/− and WT ventricular myocytes using the macropatch cell-attached mode of the patch-clamp technique, which allows for regional assessment of INa amplitude and gating properties at a physiological temperature (36°C) (17). At the lateral membrane, the number of functional sodium channels (INa amplitude) was similar in WT and CAR+/− cardiomyocytes, (Figure 6C). However, a significant decrease in INa amplitude at the intercalated disc was observed in CAR+/− cardiomyocytes as compared to WT (Figure 6D). Voltage dependence of INa activation and inactivation was not significantly different between WT and CAR+/− at the lateral membrane or at the intercalated disc (Figure 6E,F). These findings indicate that CAR haploinsufficiency leads to reduced sodium channel availability secondary to a decreased INa amplitude at the ID.

Physical interaction between NaV1.5and CAR

To determine if CAR affects NaV1.5 function through a physical association, we performed co-immunoprecipitation studies. FLAG-tagged human NaV1.5 and/or human CAR was transiently overexpressed in HEK293 cells, and cell lysates were incubated with anti-FLAG affinity beads. Overexpressed CAR was not found to precipitate with anti-FLAG beads in the absence of NaV1.5 (Figure 7). In contrast, immunoblotting of eluates from HEK293 cells overexpressing both NaV1.5 and CAR revealed that CAR co-precipitates with NaV1.5 (Figure 7), indicating a direct or indirect physical interaction between the two proteins.

Figure 7. CAR interacts with NaV1.5 in vitro.

Figure 7

Open in a new tab

The left panel shows input levels of NaV1.5 and CAR in lysates of HEK293A cells over-expressing FLAG-tagged NaV1.5 (lane 1), CAR (lane 2), or FLAG-tagged NaV1.5 together with CAR (lane 3) versus untransfected cells (lane 4). The right panel (´eluate´) shows levels of NaV1.5 and CAR after elution from FLAGagarose beads. Neither overexpressed CAR (lane 2) nor endogenous CAR (lane 4) precipitated with anti-FLAG beads in the absence of NaV1.5 (lanes 6 and 9, respectively). Eluates from HEK293 cells overexpressing both NaV1.5 and CAR revealed that CAR coprecipitates with NaV1.5 (lane 8). ´E´indicates empty lane (lane 7).

DISCUSSION

Our current work follows up on the discovery by GWAS of a genetic locus on chromosome 21q21 associated with vulnerability for VF in the setting of acute MI (4). We demonstrate that in human cardiac tissue, the risk allele at this locus is associated with decreased cardiac expression of CXADR, suggesting that reduced levels of CAR predispose to VF during ischemia. Reduced CAR expression in CAR haploinsufficient mice leads to ventricular conduction slowing and an earlier onset of inducible ventricular arrhythmias during acute regional myocardial ischemia, mediated (at least in part) through a reduction of sodium current magnitude specifically at the intercalated disc. Our findings thus establish CAR as a novel modifier of cardiac conduction and arrhythmia vulnerability in myocardial ischemia and underline the usefulness of GWAS for the identification of novel determinants of complex phenotypes such as arrhythmia.

The rs2824292 SNP at the arrhythmia susceptibility locus on 21q21 is associated with reduced cardiac CXADR expression

In a recent GWAS study, we identified a susceptibility locus for ventricular fibrillation during a first acute MI at chromosome 21q21 (4). The risk-haplotype at this locus, tagged by rs2824292, occurred at an inter-genic region and thus likely modulates risk for VF through effects on gene expression (18). Of the two genes at this locus (CXADR and BTG3), CXADR appears the most biologically plausible as previous studies have demonstrated its myocardial localization at the intercalated disc and its modulatory effect on connexin expression and atrio-ventricular conduction (9, 10). Our finding that carriership of the risk (G-) allele at rs2824292 is associated with reduced expression of the CXADR gene encoding CAR in human heart samples supports the proposition that the effect of the 21q21 locus on ischemia-induced arrhythmia is mediated by CAR and suggests that decreased cardiac levels of CAR predispose to VF. The genome-wide association study for ischemia-induced VF, which identified the chr21q21 risk locus is so far the only study for this specific clinical presentation. Genetic studies on this phenotype are hindered by the unavailability of large patient sets; indeed a study by Bugert and co-workers (19) in a small case-control set did not detect the association of rs2824292 with VF. Thus, although the effect of this variant/haplotype that we here detect on CXADR expression is in line with our previous GWAS findings, the robustness of this signal must be further validated in future association studies in larger patient sets.

Mechanism(s) of ventricular conduction slowing and arrhythmia vulnerability in CAR+/− mice

Conduction slowing in CAR+/− hearts may stem from altered cardiomyocyte excitability (i.e. sodium channel function), cell-to-cell coupling, tissue architecture, and combinations thereof. We found no evidence for myocardial hypertrophy or fibrosis in CAR+/− hearts, making electrical remodeling secondary to structural abnormalities unlikely. NaV1.5 protein levels were unchanged in CAR+/− hearts, but decreased AP upstroke velocity was observed in CAR+/− myocytes. Moreover prolongation of ECG conduction indices after flecainide administration was more pronounced in CAR+/− mice, further underlining the functional relevance of the reduced sodium channel availability secondary to CAR haploinsufficiency. In cardiac-specific inducible CAR knockout mice (i.e. post-natal CAR knockout), Cx43 protein levels in ventricular myocardium are reduced by approximately 40% (9). Although we did not detect significant reduction in Cx43 expression in CAR+/− hearts, this does not preclude the possibility that subtle alterations in Cx43 expression and/or function may be of relevance for conduction, especially during perturbed conditions such as ischemia as discussed below.

Decreased membrane excitability and electrical uncoupling are key mechanisms for conduction slowing and conduction block during ischemia and provide a pro-arrhythmic substrate for the development of re-entrant ventricular arrhythmias (20, 21). The ventricular conduction slowing in CAR+/− hearts likely exacerbates conduction slowing during ischemia, thereby contributing to arrhythmogenesis in this setting. The degree of conduction slowing observed at baseline in CAR+/− hearts was modest and not sufficient to facilitate spontaneous or induced ventricular arrhythmias. Thus, in order for (re-entrant) arrhythmias to occur, additional electrical and/or structural perturbations appear necessary, such as intercellular uncoupling or ischemia. Infusion of carbenoxolone, a gap junctional uncoupler which does not affect action potential characteristics or underlying ion currents (13), further decreased conduction velocity in both the longitudinal and the transversal direction, and increased arrhythmia susceptibility in CAR+/− but not in wild-type hearts. Of note, carbenoxolone also decreased conduction velocity in wild-type hearts, and although the magnitude of the effects of carbenoxolone was similar in both wild-type and CAR+/− hearts, the absolute values of both CVL and CVT were lowest in CAR+/− hearts after carbenoxolone. The concomitant reduction in sodium channel availability in CAR+/− hearts (but not in wild-type hearts) makes these hearts more susceptible to arrhythmias in the presence of the uncoupler carbenoxolone. Interestingly, an earlier onset of ventricular arrhythmias during ex vivo regional ischemia was also observed in hearts from Cx43 haploinsufficient mice (22), similar to our current observation in CAR+/− hearts. As carbenoxolone infusion is arrhythmogenic in CAR+/− mice, subtle alterations in gap junction proteins (undetected in our study) may contribute to the observed ischemia-induced proarrhythmia in CAR+/− hearts. On the other hand, arrhythmias in the early phase of ischemia are predominantly associated with conduction delay (which can facilitate reentry) secondary to decreased membrane excitability in depolarized tissue. Thus, reduced sodium channel availability secondary to decreased INa (as demonstrated in CAR+/− myocytes) may also contribute to increased arrhythmia susceptibility.

Several animal models have been developed to study CAR. Of these the complete (germline) constitutive knockout results in early embryonic lethality characterized by severe cardiac morphologic defects (23), cardiac-specific knockout (CAR-cKO) mice develop cardiomyopathy and/or atrio-ventricular conduction disturbance (26, 12) and the cardiac-specific inducible knockout develops progressive atrio-ventricular block (9). We chose to investigate the effects of CAR in heterozygous knockout mice in order to assess the impact of a less drastic reduction in CAR in the absence of remodeling, which may better reflect the variability in CAR levels occurring in human health and disease. Our observation that CAR haploinsufficiency results in ventricular conduction slowing appears in contrast with findings in CAR-cKO hearts, in which atrio-ventricular conduction was abnormal whereas ventricular conduction velocities remained unaffected (10). However, this discrepancy may be attributed to differences in measuring techniques, targeting strategies or genetic backgrounds.(25)

CAR affects cardiac sodium channels at the intercalated disc

CAR is a cell adhesion molecule with multiple binding partners, including the scaffolding protein ZO-1 which in turn interacts with Cx43 (8, 26). Sodium channels interact with Cx43 and are regulated by cell adhesion molecules, including the sodium channel β-subunit and ankyrin-G (27, 28). Thus, CAR may exert its effects on sodium channel availability by both direct and indirect mechanisms, which are likely not mutually exclusive. Crucially, INa was reduced at the intercalated disc of CAR+/− myocytes, whereas INa at the lateral myocyte membrane was unaffected. Since sodium channels at the intercalated disc are considered essential for proper conduction (15), a decrease in INa in this region is likely to impair function. Indeed, our findings demonstrate that a relatively modest reduction in INa at the intercalated disc is sufficient for sodium channel availability to be decreased. Moreover, we found that CAR interacts with NaV1.5 in vitro, suggesting that CAR affects sodium channel function through a physical interaction with NaV1.5. The observation that CAR impacts on INa specifically at the intercalated disc is in line with the emerging concept that cross-talk exists between structural and electrical components within the intercalated disc region (29–31).

Potential clinical implications

Our expression studies in human heart combined with electrophysiological studies in CAR+/− mice establish a novel role for CAR in mediating ventricular conduction and arrhythmogenesis during conditions such as gap junctional uncoupling or myocardial ischemia. CAR may have similar modulatory effects on cardiac electrical activity in other pathophysiological conditions including dilated cardiomyopathy and myocarditis, which require further investigation. The regulatory effects of CAR on sodium channel availability and cardiac conduction provide additional insight into the complex mechanisms underlying arrhythmogenesis during myocardial ischemia. Furthermore, genetic determinants of arrhythmia susceptibility during myocardial ischemia (such as CAR) may constitute future targets for risk stratification and prevention of potentially lethal ventricular arrhythmias in patients with coronary artery disease.

Supplementary Material

01

Acknowledgements

This study was supported by the Netherlands Heart Foundation (NHS 2005T025, 2007B010, 2010B175), the Deutsche Forschungsgemeinschaft (to M.G.), the National Institute of Health (NIH grant HL068880 to A.L.G.), the Netherlands Heart Institute (ICIN 061.02), and the Division for Earth and Life Sciences (ALW; project 836.09.003 to C.A.R.) with financial aid from the Netherlands Organization for Scientific Research (NWO). This research was performed within the framework of CTMM, the Center for Translational Molecular Medicine (www.ctmm.nl), project COHFAR (grant 01C-203). We thank Drs. J.M. Ruijter and M.W.T. Tanck for support with statistical analysis, R. Wolswinkel, L. Beekman, M. Westerveld and B. Goldbrich for technical support, and U. Wrackmeyer, U. Lisewski, M. Radke, and Y. Shi for support with animal maintenance and phenotyping. Dr. A. Varró (University of Szeged, Hungary), Dr. C.Q. Simmons (Vanderbilt University, Nashville, USA), Dr. A. Li (University of Sydney, Australia) and Drs. T. Zhang and V. Otero (University of Miami, USA) are gratefully acknowledged for assistance in obtaining and preparing human cardiac tissue samples. We thank Dr. H.L. Tan for critical reading of the manuscript.

Abbreviations and acronyms

AP

action potential

CAR

Coxsackie and adenovirus receptor

CVL

longitudinal conduction velocity

CVT

transversal conduction velocity

INa

sodium current

LAD

left anterior descending coronary artery

LV

left ventricle

MI

myocardial infarction

SNP

single nucleotide polymorphism

VF

ventricular fibrillation

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest statement

The authors declare no conflicts of interest.

References

  • 1.Rubart M, Zipes DP. Mechanisms of sudden cardiac death. J Clin Invest. 2005;115 doi: 10.1172/JCI26381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Dekker LRC, Bezzina CR, Henriques JPS, et al. Familial sudden death is an important risk factor for primary ventricular fibrillation: a case-control study in acute myocardial infarction patients. Circulation. 2006;114:1140–1145. doi: 10.1161/CIRCULATIONAHA.105.606145. [DOI] [PubMed] [Google Scholar]
  • 3.Kaikkonen KS, Kortelainen M-L, Linna E, Huikuri HV. Family history and the risk of sudden cardiac death as a manifestation of an acute coronary event. Circulation. 2006;114:1462–1467. doi: 10.1161/CIRCULATIONAHA.106.624593. [DOI] [PubMed] [Google Scholar]
  • 4.Bezzina CR, Pazoki R, Bardai A, et al. Genome-wide association study identifies a susceptibility locus at 21q21 for ventricular fibrillation in acute myocardial infarction. Nat Gen. 2010;42:688–691. doi: 10.1038/ng.623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Winkler GS. The mammalian anti-proliferative BTG/Tob protein family. J Cell Pysiol. 2010;222:66–72. doi: 10.1002/jcp.21919. [DOI] [PubMed] [Google Scholar]
  • 6.Noutsias M, Fechner H, Jonge H De, et al. Human coxsackie-adenovirus receptor is colocalized with integrins αvβ3 and αvβ5 on the cardiomyocyte sarcolemma and upregulated in dilated cardiomyopathy: implications for cardiotropic viral infections. Circulation. 2001;104:275–280. doi: 10.1161/01.cir.104.3.275. [DOI] [PubMed] [Google Scholar]
  • 7.Shaw CA, Holland PC, Sinnreich M, et al. Isoform-specific expression of the Coxsackie and adenovirus receptor (CAR) in neuromuscular junction and cardiac intercalated discs. BMC Cell Biology. 2004;5:42. doi: 10.1186/1471-2121-5-42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Cohen CJ, Shieh JT, Pickles RJ, Okegawa T, Hsieh JT, Bergelson JM. The coxsackievirus and adenovirus receptor is a transmembrane component of the tight junction. PNAS. 2001;98:15191–15196. doi: 10.1073/pnas.261452898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Lisewski U, Shi Y, Wrackmeyer U, et al. The tight junction protein CAR regulates cardiac conduction and cell-cell communication. J Exp Med. 2008;205:2369–2379. doi: 10.1084/jem.20080897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Lim B, Xiong D, Dorner A, et al. Coxsackievirus and adenovirus receptor ( CAR ) mediates atrioventricular-node function and connexin 45 localization in the murine heart. J Clin Invest. 2008;118:2758–2770. doi: 10.1172/JCI34777. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Fechner H, Noutsias M, Tschoepe C, et al. Induction of coxsackievirus-adenovirusreceptor expression during myocardial tissue formation and remodeling: identification of a cell-to-cell contact-dependent regulatory mechanism. Circulation. 2003;107:876–882. doi: 10.1161/01.cir.0000050150.27478.c5. [DOI] [PubMed] [Google Scholar]
  • 12.Shi Y, Chen C, Lisewski U, et al. Cardiac deletion of the Coxsackievirus-adenovirus receptor abolishes Coxsackievirus B3 infection and prevents myocarditis in vivo. J Am Coll Cardiol. 2009;53:1219–1226. doi: 10.1016/j.jacc.2008.10.064. [DOI] [PubMed] [Google Scholar]
  • 13.De Groot JR, Veenstra T, Verkerk AO, et al. Conduction slowing by the gap junctional uncoupler carbenoxolone. Cardiovasc Res. 2003;60:288–297. doi: 10.1016/j.cardiores.2003.07.004. [DOI] [PubMed] [Google Scholar]
  • 14.Berecki G, Wilders R, De Jonge B, Van Ginneken ACG, Verkerk AO. Re-evaluation of the action potential upstroke velocity as a measure of the Na+ current in cardiac myocytes at physiological conditions. PloS One. 2010;5:e15772. doi: 10.1371/journal.pone.0015772. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kucera JP, Rohr S, Rudy Y. Localization of sodium channels in intercalated disks modulates cardiac conduction. Circ Res. 2002;91:1176–1182. doi: 10.1161/01.res.0000046237.54156.0a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Petitprez S, Zmoos A-F, Ogrodnik J, et al. SAP97 and dystrophin macromolecular complexes determine two pools of cardiac sodium channels Nav1.5 in cardiomyocytes. Circ Res. 2011;108:294–304. doi: 10.1161/CIRCRESAHA.110.228312. [DOI] [PubMed] [Google Scholar]
  • 17.Verkerk AO, Van Ginneken ACG, Van Veen TAB, Tan HL. Effects of heart failure on brain-type Na+ channels in rabbit ventricular myocytes. Europace. 2007;9:571–577. doi: 10.1093/europace/eum121. [DOI] [PubMed] [Google Scholar]
  • 18.Cookson W, Liang L, Abecasis G, Moffatt M, Lathrop M. Mapping complex disease traits with global gene expression. Nat Rev Genet. 2009;10:184–194. doi: 10.1038/nrg2537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Bugert P, Elmas E, Stach K, et al. No evidence for an association between the rs2824292 variant at chromosome 21q21 and ventricular fibrillation during acute myocardial infarction in a German population. Clin Chem Lab Med. 2011;49:1237–1239. doi: 10.1515/CCLM.2011.190. [DOI] [PubMed] [Google Scholar]
  • 20.Janse MJ, Wit AL. Electrophysiological mechanisms of ventricular arrhythmias resulting from myocardial ischemia and infarction. Physiol Rev. 1989;69:1049–1169. doi: 10.1152/physrev.1989.69.4.1049. [DOI] [PubMed] [Google Scholar]
  • 21.Coronel R, Wilms-Schopman FJ, Opthof T, Van Capelle FJ, Janse MJ. Injury current and gradients of diastolic stimulation threshold, TQ potential, and extracellular potassium concentration during acute regional ischemia in the isolated perfused pig heart. Circ Res. 1991;68:1241–1249. doi: 10.1161/01.res.68.5.1241. [DOI] [PubMed] [Google Scholar]
  • 22.Lerner DL, Yamada KA, Schuessler RB, Saffitz JE. Accelerated onset and increased incidence of ventricular arrhythmias induced by ischemia in Cx43-deficient mice. Circulation. 2000;101:547–552. doi: 10.1161/01.cir.101.5.547. [DOI] [PubMed] [Google Scholar]
  • 23.Dorner AA, Wegmann F, Butz S, et al. Coxsackievirus-adenovirus receptor (CAR) is essential for early embryonic cardiac development. J Cell Science. 2005;118:3509–3521. doi: 10.1242/jcs.02476. [DOI] [PubMed] [Google Scholar]
  • 24.Chen J-W, Zhou B, Yu Q-C, et al. Cardiomyocyte-specific deletion of the coxsackievirus and adenovirus receptor results in hyperplasia of the embryonic left ventricle and abnormalities of sinuatrial valves. Circ Res. 2006;98:923–930. doi: 10.1161/01.RES.0000218041.41932.e3. [DOI] [PubMed] [Google Scholar]
  • 25.Fischer R, Poller W, Schultheiss H-P, Gotthardt M. CAR-diology--a virus receptor in the healthy and diseased heart. J Mol Med (Berl) 2009;87:879–884. doi: 10.1007/s00109-009-0489-5. [DOI] [PubMed] [Google Scholar]
  • 26.Giepmans BN, Moolenaar WH. The gap junction protein connexin43 interacts with the second PDZ domain of the zona occludens-1 protein. Curr Biol. 1998;8:931–934. doi: 10.1016/s0960-9822(07)00375-2. [DOI] [PubMed] [Google Scholar]
  • 27.Malhotra JD, Thyagarajan V, Chen C, Isom LL. Tyrosine-phosphorylated and nonphosphorylated sodium channel beta1 subunits are differentially localized in cardiac myocytes. J Biol Chem. 2004;279:40748–40754. doi: 10.1074/jbc.M407243200. [DOI] [PubMed] [Google Scholar]
  • 28.Mohler PJ, Rivolta I, Napolitano C, et al. Nav1.5 E1053K mutation causing Brugada syndrome blocks binding to ankyrin-G and expression of Nav1.5 on the surface of cardiomyocytes. PNAS. 2004;101:17533–11538. doi: 10.1073/pnas.0403711101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Sato PY, Musa H, Coombs W, et al. Loss of plakophilin-2 expression leads to decreased sodium current and slower conduction velocity in cultured cardiac myocytes. Circ Res. 2009;105:523–526. doi: 10.1161/CIRCRESAHA.109.201418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sato PY, Coombs W, Lin X, et al. Interactions between ankyrin-G, Plakophilin-2, and Connexin43 at the cardiac intercalated disc. Circ Res. 2011;109:193–201. doi: 10.1161/CIRCRESAHA.111.247023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Rizzo S, Lodder EM, Verkerk AO, et al. Intercalated disc abnormalities, reduced Na+ current density, and conduction slowing in desmoglein-2 mutant mice prior to cardiomyopathic changes. Cardiovasc Res. 2012;95:409–418. doi: 10.1093/cvr/cvs219. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS542095-supplement-01.pdf (605.8KB, pdf)

RESOURCES