Calmodulin Kinase II and Arrhythmias in a Mouse Model of Cardiac Hypertrophy
Background— Calmodulin kinase (CaMK) II is linked to arrhythmia mechanisms in cellular models where repolarization is prolonged. CaMKII upregulation and prolonged repolarization are general features of cardiomyopathy, but the role of CaMKII in arrhythmias in cardiomyopathy is unknown.
Methods and Results— We studied a mouse model of cardiac hypertrophy attributable to transgenic (TG) overexpression of a constitutively active form of CaMKIV that also has increased endogenous CaMKII activity. ECG-telemetered TG mice had significantly more arrhythmias than wild-type (WT) littermate controls at baseline, and arrhythmias were additionally increased by isoproterenol. Arrhythmias were significantly suppressed by an inhibitory agent targeting endogenous CaMKII. TG mice had longer QT intervals and action potential durations than WT mice, and TG cardiomyocytes had frequent early afterdepolarizations (EADs), a hypothesized mechanism for triggering arrhythmias. EADs were absent in WT cells before and after isoproterenol, whereas EAD frequency was unaffected by isoproterenol in TG mice. L-type Ca2+ channels (LTTCs) can activate EADs, and LTCC opening probability (Po) was significantly higher in TG than WT cardiomyocytes before and after isoproterenol. A CaMKII inhibitory peptide equalized TG and WT LTCC Po and eliminated EADs, whereas a peptide antagonist of the Na+/Ca2+ exchanger current, also hypothesized to support EADs, was ineffective.
Conclusions— These findings support the hypothesis that CaMKII is a proarrhythmic signaling molecule in cardiac hypertrophy in vivo. Cellular studies point to EADs as a triggering mechanism for arrhythmias but suggest that the increase in arrhythmias after β-adrenergic stimulation is independent of enhanced EAD frequency.
Received May 7, 2002; revision received June 11, 2002; accepted June 11, 2002.
Cardiac arrhythmias are a leading cause of death in patients with cardiac hypertrophy, but identification of molecular signaling pathways linking cardiac hypertrophy to disordered electrical excitability has been lacking. Calmodulin kinase (CaMK) types II and IV are both present in the nucleus,1 where they activate signaling pathways for cardiac hypertrophy,2–4⇓⇓ but CaMKII is also a proarrhythmic signaling molecule in cellular models of drug-induced action potential prolongation.5,6⇓ Upregulation of CaMKII activity and expression seems to be a general feature of cardiomyopathy from diverse causes in patients7 and animal models, 8,9⇓ suggesting the hypothesis that CaMKII is a proarrhythmic signaling molecule in cardiomyopathy where cardiac repolarization is prolonged because of electrical remodeling.
See p 1180
Reduction in repolarizing outward K+ currents and action potential and QT interval prolongation are also consistent findings in electrical remodeling in cardiac hypertrophy. However, the electrical remodeling phenotype, per se, does not explain the mechanism of arrhythmias in cardiac hypertrophy. On the other hand, L-type Ca2+ current (ICa) and Na+/Ca2+ exchanger current (INa/Caex) are candidate inward currents for initiating arrhythmia-triggering afterdepolarizations in electrically remodeled myocardium.10,11⇓ Whereas CaMKII is hypothesized to favor afterdepolarizations attributable to ICa 5 and INa/Caex 6 in cellular models, CaMKII has not been demonstrated to be a proarrhythmic signal in vivo.
We developed a mouse model of cardiac hypertrophy using transgenic (TG) expression of a constitutively active form of CaMKIV, where endogenous CaMKII activity can be readily distinguished from TG CaMKIV activity using specific inhibitors and substrates. These mice have cardiac hypertrophy and reduced systolic function2 and significantly increased CaMKII activity and expression. We studied these mice to test the hypothesis that CaMKII can be a proarrhythmic signaling molecule in vivo. In this study, we report in vivo, cellular, and biochemical evidence that CaMKII is a critical signal for arrhythmias in cardiac hypertrophy.
CaMKIV Transgenic Mice
TG mice expressed a truncated, constitutively active form of human CaMKIV lacking the calmodulin-binding domain.2 TG mice develop moderate cardiac hypertrophy and ≈50% reduction in left ventricular ejection fraction. Experiments were performed on 8- to 24-week-old TG mice and wild-type (WT) littermate controls.
Mice were anesthetized (pentobarbital 33 μg/g and ketamine 33 μg/g IP) before placing a transmitter (Data Sciences International) into the abdominal cavity with subcutaneous electrodes in a lead I configuration. ECG intervals were determined in ambulatory, unanesthetized mice by signal averaging 10-second epochs every 5 minutes for a 30-minute baseline period. Continuous recording was performed for 30-minute intervals during arrhythmia screening. Signal-averaging was performed using custom software developed at Vanderbilt University. Interval measurements were performed without explicit knowledge of genotype, but QT interval prolongation was always apparent in the TG mice. QT intervals were corrected (QTc) for heart rate by a formula developed for mice.12
Arrhythmia Induction and Screening
Arrhythmias were categorized into 5 groups and assigned the following point values: no arrhythmias, 0 points; premature atrial or ventricular beats, 1 point; supraventricular tachycardia or paired premature ventricular beats, 2 points; bigeminal or trigeminal premature ventricular beats or nonsustained ventricular tachycardia (≥3 consecutive premature ventricular beats), 3 points; and sustained ventricular tachycardia (>10 consecutive premature ventricular beats) or polymorphic ventricular tachycardia, 4 points.
Mice were treated with the CaMK inhibitory agent KN-93 (10 to 30 μmol/kg IP) or the inactive congener KN-92 (30 μmol/kg IP)13 10 minutes before isoproterenol (100 μg IP). KN-93 and KN-92 were tested in the same mouse on different days. ECGs were obtained from unanesthetized and unrestrained mice 30 minutes after isoproterenol. Recordings were analyzed offline and coded for arrhythmias.
CaMKII Activity and Expression
CaMKII activity was determined in the presence of Ca2+/CaM from fresh ventricular homogenates5,14⇓ using syntide 2, a synthetic substrate with ≈50-fold selectivity for CaMKII over CaMKIV.15 Ventricular extracts (30 μg protein per lane) were analyzed by SDS-PAGE and immunoblotted using a CaMKII antibody that specifically recognizes the δ isoform (a generous gift from Dr H. Singer, SUNY, Albany, NY) and a control PP1β-specific antibody.16 Blots were developed using colorimetric reagents with alkaline phosphatase conjugated secondary antibodies and digitized images were quantified using NIM Image.
The CaMKII inhibitory peptide AC3-I (KKALHRQEAVDCL, IC50 ≈3 μmol/L)17 (Macromolecular Resources) is a modified CaMKII substrate; AC3-C (KKALHAQERVDCL) is an inactive control peptide (IC50 >500 μmol/L). AC3-I (40 to 100 μmol/L) and AC3-C (100 μmol/L) were dialyzed into cells for 5 to 10 minutes before experiments. CaMKII inhibitory peptides were engineered for cell membrane permeability with separate minigenes encoding AC3-I and AC3-C using pGEX-3X-MTS2 (a generous gift from Drs Rojas and Lin, Vanderbilt University, Nashville, Tenn), as described.18 Cardiomyocytes were exposed to cell membrane permeant peptides (1 μmol/L) for ≥30 minutes before the experiments.
The INa/Caex inhibitory peptide (XIP, RREIFYKYVYKRYRAGKQRF)19 and the inactive control peptide scrambled XIP (sXIP) (Macromolecular Resources) were dialyzed for 5 to 10 minutes (10 μmol/L) before initiating experiments.
Myocyte Isolation and Electrophysiology
Ventricular myocytes were isolated as previously described.20
Whole-cell mode voltage clamp measured transient (Ito) and sustained (Isus) components of repolarizing K+ current (T=34°C to 36°C). Isus was the residual current at the end of a 450-ms depolarizing pulse (0.33 Hz), and Ito was the difference between peak outward K+ current and Isus (Figure 1e).21 The bath solution for K+ current voltage clamp studies was (in mmol/L) N-methyl-d-glucamine 149, HEPES 5, glucose 5, KCl 1, and MgCl2 5, and the pH was adjusted to 7.4 with 12 N HCl.14 The pipette solution was (in mmol/L) K aspirate 120, HEPES 5, KCl 25, Na2ATP 4, MgCl2 1, Na2 phosphocreatine 2, NaGTP 2, CaCl2 1, and EGTA 10, and the pH was adjusted to 7.2 with 1 N KOH. On-cell mode voltage clamp configuration was used to measure single LTCC currents, using Ba2+ or Ca2+ (both 110 mmol/L) as charge carrier, as previously described by us.20 β-Adrenergic signaling was activated for LTCC recordings using isoproterenol (2 μmol/L) and isobutylmethylxanthine (20 μmol/L).22 Current clamp was used for stimulating action potentials (0.5 Hz) in physiological solutions (T=34°C to 36°C). Action potential duration was measured at 50% (APD50) and 90% (APD90) repolarization to baseline. The bath solution contained (in mmol/L) NaCl 140, HEPES 5, glucose 10, KCl 5.4, CaCl2 2.5, and MgCl2 1, and the pH was adjusted to 7.4 with 10 N NaOH. The pipette (intracellular) solution was the same as listed above for K+ current experiments. Junction potentials between pipette and bath solutions were compensated electronically. Early afterdepolarizations (EADs) were defined as discrete oscillations in repolarization during the action potential plateau.14
The null hypothesis was rejected for P<0.05 using Student’s unpaired t test or ANOVA, as appropriate. The Wilcoxon ranked sign test was used for comparison of KN-93 and KN-92 effects on arrhythmias scores, and Fisher’s exact test was used to compare the frequency of EADs between WT and TG mice. Data were expressed as mean±SEM.
Electrical Remodeling in CaMKIV TG Mice
CaMKIV TG mice have increased QT intervals (Figures 1a and 1b), prolonged action potential durations (APDs) (Figures 1c and 1d), and reduced repolarizing K+ currents (Figure 1e). These findings show that the CaMKIV TG mouse has important electrical changes seen generally in cardiomyopathy and suggest that it may be a useful new model to probe the molecular signaling mechanisms for arrhythmias in cardiac hypertrophy.
CaMKII Expression and Activity Are Increased in TG Hearts
CaMKII activity and expression are both increased in CaMKIV TG mice (Figure 2), as occurs in humans and other animal models of cardiomyopathy.7–9⇓⇓ Based on the combined presence of electrical remodeling (Figure 1) and increased CaMKII activity (Figure 2), we tested whether CaMKIV TG mice had arrhythmias and if these arrhythmias could be suppressed by CaMKII inhibition.
Increased Arrhythmias in TG Mice Are Suppressed by Inhibition of CaMK
Although one CaMKIV TG mouse had spontaneous torsade de pointes (Figure 3a), a form of ventricular tachycardia linked to Ca2+/calmodulin (CaM)-dependent signaling,23,24⇓ and arrhythmias were significantly more common in TG than WT mice (Figure 3c), arrhythmia scores were low under basal conditions. Isoproterenol was administered to increase arrhythmia scores, based on the reasoning that β-adrenergic agonists are known to increase [Ca2+]i,25–27⇓⇓ activate CaMKII, 26 and favor arrhythmia-triggering afterdepolarizations.25,27⇓ Arrhythmias were observed significantly more frequently in unanesthetized and unrestrained TG mice at baseline, and high-grade arrhythmias (point score ≥2) occurred frequently in TG mice after isoproterenol (Figure 3), consistent with the increased tendency for arrhythmias in electrically remodeled myocardium.
To test the hypothesis that the enhanced CaMKII activity (Figure 2) contributed to high-grade arrhythmias that were frequent in TG mice after isoproterenol, TG mice were pretreated with the CaMK inhibitory agent KN-93 or the inactive congener KN-92. KN-93 selectively targets endogenous CaMK, because it acts as a competitive and noncompetitive inhibitor for CaM binding,13 and the CaM binding domain is absent on the TG CaMKIV.2 These paired experiments showed KN-93 significantly reduced arrhythmia severity in TG mice (Figure 3e) but not in isoproterenol-treated WT mice (not shown), supporting the concept that CaMKII is a proarrhythmic molecule in this model of cardiac hypertrophy.
CaMKII Activity Is Required for EADs in CaMKIV TG Mice
EADs are an important trigger for arrhythmias in electrically remodeled myocardium,10,11⇓ and EADs were only observed in ventricular myocytes from TG mice at baseline (Figures 4a and 4c) and after isoproterenol (Figure 4d). Interestingly, EAD frequency was not affected by isoproterenol (P=0.46 compared with baseline for TG cells). Isoproterenol induced complex effects on APD, lengthening APD50 in TG and WT cells but shortening APD90 only in TG cardiomyocytes (Figure 4b). The findings up to this point show that TG mice have increased arrhythmias and frequent EADs and suggest that isoproterenol may enable EADs to more effectively trigger arrhythmias without increasing overall EAD frequency.
To examine the hypothesized cellular basis for the link between increased CaMKII activity and arrhythmias, we measured the response of EADs to the CaMKII inhibitory peptide AC3-I.17 CaMKII activity was required for EADs because EADs were prevented by AC3-I but not by the inactive control peptide AC3-C (Figure 4e). Both LTCC20 and the Na+/Ca2+ exchanger6 cause inward currents regulated by CaMKII activity that may initiate EADs. However, a INa/Caex inhibitory peptide did not prevent EADs (Figure 4e), indicating that INa/Caex did not cause EADs in this model. EAD initiation was within the ICa window potential range (−27.5±0.2 mV for 534 EADs from 9 cells), suggesting that LTCC activity could be responsible for EADs seen in these cells.28
Action Potential Prolongation Alone Does Not Cause EADs
EADs can be prevented by shortening action potential repolarization,29 so we measured APD in isolated TG myocytes to determine if the mechanism of CaMKII inhibition in suppressing EADs was related to shortening repolarization. CaMKII inhibition did not shorten APD (Figure 5a). In contrast, the INa/Caex inhibitory peptide XIP shortened APD in WT but not TG ventricular myocytes (Figure 5c), indicating that XIP produced a significant effect at the concentration used in these experiments and suggesting that INa/Caex was not a critical determinant of the action potential prolongation seen in the electrically remodeled TG cardiomyocytes. Taken together, these findings indicate that CaMKII activation, and not APD prolongation alone, is required for EADs and that these EADs are attributable to a INa/Caex-independent conductance.
Increased LTCC Activity in TG Mice Is CaMKII Dependent
LTCC activity is increased in myopathic human hearts,30 and CaMKII increases LTCC opening probability (Po) in normal adult cardiomyocytes,20 suggesting the possibility that increased CaMKII activity in TG mouse hearts (Figure 2) and diseased human hearts7 could drive increased LTCC activity. LTCC Po was higher in TG than WT cells at baseline (Figures 6a and 6c) and after isoproterenol (Figure 6d). To determine if increased CaMKII activity (Figure 2) was the molecular mechanism for increased LTCC Po in TG mice, we developed a cell membrane permeant form of AC3-I (Figure 6b). AC3-I, but not a cell membrane permeant control peptide, eliminated the increased LTCC Po present in TG mice (Figure 6c). These findings indicate that CaMKII activity is the cause of increased LTCC activity seen in the TG mice at baseline and suggest a mechanistic framework for understanding the efficacy of acute suppression of CaMKII in reducing arrhythmias (Figure 3) and EADs (Figure 4).
CaMK and the Arrhythmogenic Phenotype in Cardiomyopathy
CaMK types I,2 II, 3,33⇓ and IV2 are all implicated in cardiac hypertrophy, but CaMKII is by far the most abundant CaM-activated kinase in heart. Upregulation of CaMKII seems to be a general feature of cardiomyopathy in humans,7 rats, 8 and mice (Figure 2), although the mechanism for CaMKII upregulation is presently unknown. CaMKII targets key control molecules for intracellular Ca2+ homeostasis in cardiomyocytes, and increased CaMKII may trigger arrhythmia-initiating afterdepolarizations5 by activating LTCCs20 or increasing INa/Caex during action potential prolongation.6 The present results support the hypothesis that CaMKII is a critical molecular signal for arrhythmias in cardiac hypertrophy in vivo and suggest that increased LTCC Po and EADs may be CaMKII-driven components of the cellular arrhythmia mechanism.
The present findings support the concept that the increased CaMKII present in cardiomyopathy is arrhythmogenic and that CaMKII-dependent arrhythmias are additionally enhanced after activation through β-adrenergic stimulation, perhaps by virtue of the increased [Ca2+]i that follows generation of protein kinase A.26 On the other hand, β-adrenergic receptor blockade significantly reduces sudden death in patients with cardiac hypertrophy and heart failure,31 raising the possibility that part of the salutary effect of β-adrenergic receptor antagonist drugs may be mediated through secondary actions on CaMKII signaling.
Arrhythmia Mechanisms in Cardiomyopathy
There is an increasing recognition that arrhythmia mechanisms in cardiomyopathy involve both cellular and tissue remodeling. Cellular studies consistently reveal APD prolongation and an increased tendency for afterdepolarizations that are a hypothesized focal mechanism for arrhythmia triggering.10,11⇓ Focal cellular mechanisms seem to be important for arrhythmia initiation in patients with cardiomyopathy,32 but changes in the myocardium, involving intercellular coupling, scaring, and fibrosis, also constitute a macroscopic arrhythmogenic substrate that contributes to arrhythmia maintenance. The TG mice have more arrhythmias and frequent EADs at baseline, but the finding that EAD frequency is unchanged whereas arrhythmias increase after isoproterenol suggests that the proarrhythmic action of isoproterenol occurs at the tissue level, perhaps by enhancing EAD propagation. The present studies are the first to implicate CaMKII-dependent signaling in arrhythmias in cardiac hypertrophy. Additional studies will be required to address the effects of CaMKII on macroscopic arrhythmia mechanisms.
This work was supported by National Heart, Lung, and Blood Institute grants HL03727 and HL62494 (to Dr Anderson). Dr Anderson is an Established Investigator of the American Heart Association and a Stahlman Scholar. We thank Martha Bass and Jinying Yang for technical assistance and Drs Jeffrey R. Balser and Lou DeFelice for helpful comments.
↵*These authors contributed equally to this study.
- ↵Ramirez MT, Zhao XL, Schulman H, et al. The nuclear δB isoform of Ca2+/calmodulin-dependent protein kinase II regulates atrial natriuretic factor gene expression in ventricular myocytes. J Biol Chem. 1997; 272: 31203–31208.
- ↵Zhang T, Johnson EN, Gu Y, et al. The cardiac-specific nuclear δ(B) isoform of Ca2+/calmodulin-dependent protein kinase II induces hypertrophy and dilated cardiomyopathy associated with increased protein phosphatase 2A activity. J Biol Chem. 2002; 277: 1261–1267.
- ↵Wu Y Roden DM, Anderson ME. Calmodulin kinase inhibition prevents development of the arrhythmogenic transient inward current. Circ Res. 1999; 84: 906–912.
- ↵Hoch B, Meyer R, Hetzer R, et al. Identification and expression of δ-isoforms of the multifunctional Ca2+/calmodulin-dependent protein kinase in failing and nonfailing human myocardium. Circ Res. 1999; 84: 713–721.
- ↵Priebe L, Beuckelmann DJ. Simulation study of cellular electric properties in heart failure. Circ Res. 1998; 82: 1206–1223.
- ↵Anderson ME, Braun AP, WuY, et al. KN-93, an inhibitor of multifunctional Ca2+/calmodulin-dependent protein kinase, decreases early afterdepolarizations in rabbit heart. J Pharmacol Exp Ther. 1998; 287: 996–1006.
- ↵Miyano O, Kameshita I, Fujisawa H. Purification and characterization of a brain-specific multifunctional calmodulin-dependent protein kinase from rat cerebellum. J Biol Chem. 1992; 267: 1198–1203.
- ↵Li Z, Nicoll DA, Collins A, et al. Identification of a peptide inhibitor of the cardiac sarcolemmal Na+-Ca2+ exchanger. J Biol Chem. 1991; 266: 1014–1020.
- ↵duBell WH, Lederer WJ, Rogers TB. K+ currents responsible for repolarization in mouse ventricle and their modulation by FK-506 and rapamycin. Am J Physiol. 2000; 278: H886–H897.
- ↵Yue DT, Herzig S, Marban E. β-Adrenergic stimulation of calcium channels occurs by potentiation of high-activity gating modes. Proc Natl Acad Sci U S A. 1990; 87: 753–757.
- ↵Mazur A, Roden DM, Anderson ME. Systemic administration of calmodulin antagonist W-7 or protein kinase A inhibitor H-8 prevents torsade de pointes in rabbits. Circulation. 1999; 100: 2437–2442.
- ↵Gbadebo TD, Trimble RW, Khoo MSC, et al. Calmodulin inhibitor W-7 unmasks a novel electrocardiographic parameter that predicts initiation of Torsade de Pointes. Circulation. 2002; 105: 770–774.
- ↵De Ferrari GM, Viola MC, D’Amato E, et al. Distinct patterns of calcium transients during early and delayed afterdepolarizations induced by isoproterenol in ventricular myocytes. Circulation. 1995; 91: 2510–2515.
- ↵2 Hirano Y, Moscucci A, January CT. Direct measurement of L-type Ca2+ window current in heart cells. Circ Res. 1992; 70: 445–455.
- ↵Fish FA, Prakash C, Roden DM. Suppression of repolarization-related arrhythmias in vitro and in vivo by low-dose potassium channel activators. Circulation. 1990; 82: 1362–1369.
- ↵Schroder F, Handrock R, Beuckelmann DJ, et al. Increased availability and open probability of single L-type calcium channels from failing compared with nonfailing human ventricle. Circulation. 1998; 98: 969–976.
- ↵Pogwizd SM, McKenzie JP, Cain ME. Mechanisms underlying spontaneous and induced ventricular arrhythmias in patients with idiopathic dilated cardiomyopathy. Circulation. 1998; 98: 2404–2414.
- ↵Zhang T, Johnson EN, Gu Y, et al. The cardiac-specific nuclear delta (B) isoform of Ca2+/calmodulin-dependent protein kinase II induces hypertrophy and dilated cardiomyopathy associated with increased protein phosphate 2A activity. J Biol Chem. 2002; 277: 1261–1267.