Familial Hypertrophic Cardiomyopathy-Linked Mutant Troponin T Causes Stress-Induced Ventricular Tachycardia and Ca2+-Dependent Action Potential Remodeling
The cardiac troponin T (TnT) I79N mutation has been linked to familial hypertrophic cardiomyopathy and high incidence of sudden death, despite causing little or no cardiac hypertrophy in patients. Transgenic mice expressing mutant human TnT (I79N-Tg) have increased cardiac contractility, but no ventricular hypertrophy or fibrosis. Enhanced cardiac function has been associated with myofilament Ca2+ sensitization, suggesting altered cellular Ca2+ handling. In the present study, we compare cellular Ca2+ transients and electrophysiological parameters of 64 I79N-Tg and 106 control mice in isolated myocytes, isolated perfused hearts, and whole animals. Ventricular action potentials (APs) measured in isolated I79N-Tg hearts and myocytes were significantly shortened only at 70% repolarization. No significant differences were found either in L-type Ca2+ or transient outward K+ currents, but inward rectifier K+ current (IK1) was significantly decreased. More critically, Ca2+ transients of field-stimulated ventricular I79N-Tg myocytes were reduced and had slow decay kinetics, consistent with increased Ca2+ sensitivity of I79N mutant fibers. AP differences were abolished when myocytes were dialyzed with Ca2+ buffers or after the Na+-Ca2+ exchanger was blocked by Li+. At higher pacing rates or in presence of isoproterenol, diastolic Ca2+ became significantly elevated in I79N-Tg compared with control myocytes. Ventricular ectopy could be induced by isoproterenol-challenge in isolated I79N-Tg hearts and anesthetized I79N-Tg mice. Freely moving I79N-Tg mice had a higher incidence of nonsustained ventricular tachycardia (VT) during mental stress (warm air jets). We conclude that the TnT-I79N mutation causes stress-induced VT even in absence of hypertrophy and/or fibrosis, arising possibly from the combination of AP remodeling related to altered Ca2+ transients and suppression of IK1.
- familial hypertrophic cardiomyopathy
- troponin T
- ventricular tachycardia
- action potential remodeling
- Ca2+ transient
Familial hypertrophic cardiomyopathy (FHC) is an autosomal-dominant disease resulting from mutations in genes encoding cardiac contractile proteins and is an important cause of sudden cardiac death.1 Genotype/phenotype correlation studies suggest that the prognostic significance of most mutations is related to the degree of cardiac hypertrophy and fibrosis.2,3 An exception to this are patients with cardiac troponin T (TnT) mutations such as TnT-I79N,4 who often die suddenly at a young age,5 even though cardiac hypertrophy and fibrosis are either mild or nonexistent.6 Although it is generally assumed that these sudden deaths are caused be ventricular arrhythmias, such evidence has been difficult to obtain,7 and no data exist for the TnT-I79N mutation.
In vitro studies of skinned fibers reconstituted with mutant TnT protein show that FHC-linked troponin T mutations almost universally increase myofilament Ca2+ sensitivity.8 Because Ca2+ binding to the troponin complex represents the largest component of dynamic Ca2+ buffering during the cardiac cycle,9 our modeling studies predict that the increased myofilament Ca2+ sensitivity would significantly alter intracellular Ca2+ transients.10 Thus, TnT mutations that change intracellular Ca2+ handling may lead to action potential remodeling and possibly ventricular arrhythmias responsible for sudden cardiac death. To test this hypothesis, we have generated transgenic mice with cardiac-targeted expression of mutant human cardiac TnT-I79N (I79N-Tg) that show increased myofilament Ca2+ sensitivity,11 enhanced cardiac contractility, and impaired relaxation,10 but no ventricular hypertrophy or fibrosis.10 In the present study, we examine the electrophysiological and Ca2+ signaling consequences of the TnT-I79N mutation in transgenic mice. Our data suggests that the TnT-I79N mutation can cause stress-induced ventricular tachycardia even in absence of hypertrophy and/or fibrosis, arising possibly from the combination of action potential remodeling, altered cytosolic Ca2+ transients, and decreased inward rectifier K+ current.
Materials and Methods
Groups of 4- to 6-month-old mice expressing human wild-type cardiac TnT (WT-Tg), mutant cardiac TnT (I79N-Tg), and nontransgenic littermates (Non-Tg) were used for all studies, which were performed according to NIH guidelines and approved by the institutional animal care and use committee.
Telemetric ECG Recordings
Telemetric ECGs were continuously recorded from 11 mice (2 WT-Tg, 4 Non-Tg, 5 I79N-Tg) during 48 hours normal activity and during defined stress tests (see the expanded Materials and Methods section, available in the online data supplement at http://www.circresaha.org), namely: (1) 4 minutes of continuously swimming in cages filled with water (37°C); and (2) repetitive warm air stress. This “mental stress” procedure with warm air jets repetitively increased heart rate (online Figure 1) and has been reported to increase systemic blood pressure and sympathetic nerve activity in rats12 and mice.13 All stress test protocols were performed by a single operator and were performed simultaneously on matched pairs of mice.
ECG Recordings in Anesthetized Mice
Thirty-nine mice (12 WT-Tg, 13 Non-Tg, 14 I79N-Tg) were anesthetized with ketamine (100 mg/kg) and xylazine (5 mg/kg) IP, and bipolar limb lead ECGs recorded.14 After 5 minutes of continuous baseline ECG recordings, mice received either 0.1 mg/kg or 1.5 mg/kg of isoproterenol IP, followed by 15 minutes of additional ECG monitoring.
Isolated Perfused Heart Preparation
Sixty-five mice (18 WT-Tg, 23 Non-Tg, 24 I79N-Tg) were anesthetized with 20 mL/kg of 2% tribromoethanol via IP injection. Experiments were carried out either in unloaded15 or isovolumetrically-beating hearts (see online data supplement).10
Volume-Conducted ECG, Transmembrane, and Monophasic Action Potential Recordings
Volume-conducted ECG, transmembrane action potential (TAP) and monophasic action potential (MAP) recordings were obtained from isolated perfused mouse heart as previously described.15
Myocyte Isolation, Action Potential, and Voltage Clamp Measurements
Ventricular myocytes were enzymatically isolated from 20 control (12 Non-Tg and 8 WT-Tg) and 13 I79N-Tg hearts (see online data supplement).14 Ca2+-tolerant cells were used for voltage-clamp measurements of L-type Ca2+ and K+ currents and current-clamp action potential measurement using the ruptured patch method (see online data supplement), and for measurements of [Ca2+]i during field stimulation. A subset of cells was used for simultaneous membrane potential measurements using the perforated patch method (see online data supplement). Action potentials were measured after 5 minutes of steady-state pacing at 1 Hz and 5 Hz.
Measurement of Intracellular [Ca2+]
Cells were loaded with membrane-permeable fura-2 acetoxymethyl ester (fura-2AM, Molecular Probes, Inc), and intracellular Ca2+ transients were measured (see online data supplement). A subset of cells was exposed to 5 mmol/L caffeine for 5 seconds using a rapid concentration clamp system. Amplitudes of caffeine-induced Ca2+ transients were used as estimates of sarcoplasmic reticulum (SR) Ca2+ content.
ECG and Action Potential Data Analysis
All data were analyzed in blinded fashion without knowledge of the genotype. All ECG recordings were manually analyzed for heart rate and PR interval during each experimental period. The entire recording periods were scrutinized for ventricular arrhythmias (see online data supplement). Microelectrode, MAP, and single-cell AP data were analyzed using a custom-built software program written in LabVIEW (National Instruments) as previously described (see online data supplement).15
Mean±SEM values are given, unless otherwise indicated. Mean values were compared with single factor analysis of variance (ANOVA). Post hoc Student’s t test analysis was performed whenever significant differences were detected by ANOVA. Fischer’s exact test (one-tailed) was used to compare incidence of arrhythmias between groups.
To test the hypothesis that the FHC-linked TnT-I79N mutation causes ventricular arrhythmias, we compared ECG and action potential recordings from transgenic mice with cardiac-targeted expression of human troponin T containing the single amino acid I79N mutation (I79N-Tg) with two control groups: mice expressing human wild-type troponin T (WT-Tg) or no transgene (Non-Tg). In this study, we pooled the data from the two control groups, because our previous characterization did not show significant differences between ECG parameters or incidence of ventricular arrhythmias of WT-Tg and Non-Tg mice.11
Telemetric ECG Recordings in Freely Moving Mice
I79N-Tg mice (n=5) demonstrated a nonsignificant trend toward slower average heart rate and longer PR interval values compared with control mice (2 WT-Tg and 4 Non-Tg) during normal activity (heart rate, 556±21 versus 600±18 bpm; P=0.06; PR interval, 37±1.1 ms versus 33±1.1 ms; P=0.06). The average heart rate of I79N-Tg mice was significantly slower during swimming exercise (674±8 versus 740±9 bpm; P<0.05), but not during “mental stress” in form of warm air jets12 (652±14 versus 651±7 bpm; P=NS).
Analysis of 48 hours of continuous telemetric ECG recordings did not reveal any significant differences in the incidence of ventricular tachycardia (VT) between I79N-Tg and control mice (Figure 1). Similarly, no VT occurred in either group during the swimming exercise protocol (5 minutes of exercise and 55 minutes of postexercise ECG monitoring). Warm air stress, however, induced short runs of nonsustained VT in 4 out of 5 I79N-Tg mice (mean 1.8±0.7 VT episodes per mouse), compared with 1 out of 6 control mice (mean 0.2±0.2 VT episodes per mouse; P=0.045). In the hour after the warm air stress, the rate of VT was higher in I79N-Tg mice compared with control mice (4.2±0.7 VT episodes per mouse in 5/5 I79N-Tg mice versus 1.2±0.2 VT episodes per mouse in 3/6 control mice; P=0.12). One I79N-Tg mouse developed several long episodes on self-terminating VT (maximal length 29 beats; Figure 1B). None of the control mice had VT episodes lasting longer than 3 consecutive premature ventricular beats (PVBs) during any recording period analyzed.
ECG Recordings in Anesthetized Mice
Under ketamine/xylazine anesthesia, heart rate and incidence of ventricular ectopy were not significantly different between 13 I79N-Tg and 25 control mice (13 WT-Tg and 12 Non-Tg). As previously reported,10 PR interval was significantly longer in I79N-Tg compared with control mice. On exposure to isoproterenol (ISO 0.1 mg/kg IP), heart rate increased to similar values in both groups of mice (I79N-Tg 472±36 versus control 473±19 bpm; P=0.96). Compared with control mice, however, I79N-Tg mice had significantly higher rate of PVBs (inset, Figure 2A) during 15 minutes of continuous ECG monitoring after the ISO injection (Figures 2A). PVBs were present in 8 of 9 I79N-Tg mice but only in 6 of 15 control mice (P=0.02). Short runs (3 to 6 beats) of VT were observed in 2 of 9 I79N-Tg mice, but only in 1 of 15 control mice (P=0.31). Administration of higher dose of ISO (1.5 mg/kg), which initially increased the heart rate in both groups similarly (I79N-Tg 499±51 versus control 515±18 bpm; P=0.79), resulted in an even higher rate of PVBs in I79N-Tg compared with control mice (Figure 2A). In addition, there was a trend toward increased incidence of nonsustained VT (Figure 2B) in I79N-Tg mice (3 of 5), which had multiple episodes compared with rare episodes in 2 control mice (2 of 10; P=0.16). VT was always self-terminating after 3 to 25 beats. Consistent with our previous report in this model,10 I79N-Tg mice frequently developed conduction blocks and 3 mice died with complete heart block present on ECG. To test whether stress-induced hemodynamic changes or atrioventricular conduction blocks may have contributed to the high incidence of ventricular arrhythmias, Langendorff-perfused hearts were used to examine the etiology of the ventricular ectopy under controlled perfusion conditions.
ECG Recordings in Langendorff-Perfused Hearts
Spontaneously beating, isovolumically contracting isolated hearts were perfused at a constant pressure of 75 mm Hg with Krebs-Henseleit buffer containing 1.8 mmol/L Ca2+. I79N-Tg hearts had significantly more ventricular ectopy (4.8±1.4 PVB/min, n=12) compared with control hearts (1.1±0.3 PVB/min, n=24 [10 WT-Tg and 14 Non-Tg]; P<0.01). Addition of isoproterenol (100 nmol/L) to the perfusate increased heart rates to comparable levels in I79N-Tg and control mice (533±12 versus 539±11 bpm; P=NS). As was observed in vivo, several minutes after the initial increase in heart rate, a second-degree AV conduction block developed transiently in 8 of 12 I79N-Tg hearts, but only in 7 of 24 control hearts (P=0.04). The AV blocks resolved after several minutes. Complete heart block was not observed. To avoid complications secondary to AV conduction blocks, the PVB rate was quantified before the onset of AV block. Although isoproterenol increased the rate of ventricular ectopy in both groups, the rate was significantly higher in I79N-Tg (12±3.0 PVB/min) versus control hearts (5.2±1.1 PVB/min; P<0.01).
Pacing Rate and Ca2+ Dependence of Ventricular Ectopy
Because the TnT-I79N mutation increases myofilament Ca2+ sensitivity11 and impairs ventricular relaxation especially at higher extracellular [Ca2+],10 Ca2+ dependence of ventricular ectopy was examined (9 I79N-Tg and 17 control mice [8 WT-Tg and 9 Non-Tg]). Unloaded hearts were paced to avoid any complications from heart blocks and intraventricular balloon. In 1.2 mmol/L [Ca2+]o, the rate of ventricular ectopy (PVBs) was significantly different only at the highest pacing rates (Figure 3A). Episodes of nonsustained VT were not observed in either group. In 3.2 mmol/L [Ca2+]o, however, the rate of PVBs was significantly increased in I79N-Tg hearts at most pacing rates (Figure 3B). Nonsustained VT occurred in 2 of 9 I79N-Tg hearts, but in none of 16 control hearts (P=0.12). Addition of isoproterenol in the presence of higher [Ca2+]o further increased ventricular ectopy in both groups and increased the incidence of VT (8/9 I79N-TG versus 8/17 controls; P=0.04). Thus, interventions that increase [Ca2+]i (isoproterenol, fast pacing rate, higher [Ca2+]o) appear also to increase the incidence of ventricular ectopy in I79N-Tg mice.
Microelectrode and MAP Measurements of Action Potential in Intact Hearts
To investigate potential cellular mechanisms responsible for the ventricular arrhythmias, KCl-filled microelectrodes were used to measure ventricular action potentials15 in isolated perfused hearts paced at 400 bpm. In 1.2 mmol/L [Ca2+]o, action potential wave shape rather than overall action potential duration was found to be significantly altered in I79N-Tg hearts (Figure 4A): the ventricular action potential had primarily a lower terminal repolarization phase, which could be quantified as a decrease in action potential duration at 70% repolarization (APD70, 14±1.2 ms, n=8, versus 23±2.2 ms, n=10, P<0.01, in I79N-Tg and control hearts, respectively). On the other hand, resting potential, overshoot potential, APD30, APD50, and APD90 were not significantly different in the two groups of mice. Consistent with the shorter APD70, there was a trend toward shorter effective refractory period in I79N-Tg (36±3 ms) compared with control mice (45±4 ms; P=0.065).
In 1.8 mmol/L [Ca2+]o, miniaturized contact electrodes were used to record monophasic action potentials (MAPs), because it was difficult to use microelectrodes reliably for the more vigorously contracting hearts. The reliability of MAP with respect to duration and shape of microelectrode recordings was recently shown.15 Figure 4 compares microelectrode (panel A) and MAP recordings (panel B) from I79N-Tg and control hearts. In a manner similar to that found when using microelectrodes, MAP wave shape rather than overall MAP duration was changed (Figure 4B): left-ventricular epicardial MAPs of I79N-Tg hearts had significantly shorter action potentials at 50% and 70%, but not at 30% and 90% repolarization (Figure 4C). Results from RV epicardial MAP recordings also demonstrated a significantly shorter MAPD70 in I79N-Tg compared with control hearts (RV-MAPD70, I79N-Tg 16±3.2 versus control 20±5.8 ms; P<0.05; and no significant difference in MAPD30, MAPD50, and MAPD90.
Measurement of Action Potential and Membrane Currents in Isolated Ventricular Myocytes
To exclude any potential contribution of Ca2+ binding to mutant Troponin complex on action potential duration, myocytes were dialyzed with pipette solution containing high (14 mmol/L) concentration of EGTA that completely abolished contractions. Unlike the isolated heart, action potential durations measured in highly Ca2+-buffered myocytes were not statistically different between control and I79N-Tg myocytes (Figure 5A). The duration of action potentials recorded from ventricular myocytes were on average significantly shorter than those recorded either in the isolated heart (compare Figure 4) or in myocytes using the perforated patch technique (compare Figure 6).
Voltage-clamp measurements of L-type Ca2+ current (Figure 5B) and depolarization-activated outward K+ currents (Figure 5C) also failed to demonstrate significant differences between control and I79N-Tg myocytes. The outward component of the inward rectifier K+ current (IK1) was modestly, but significantly reduced in I79N-Tg compared with control myocytes (Figure 5D, inset). Because reduction of IK1 would tend to prolong repolarization, these data suggest that the shorter APD70 of I79N-Tg hearts is unlikely to result from changes in Ca2+ or K+ membrane currents, but may depend on the characteristics of intracellular Ca2+ signaling.
Simultaneous Measurement of Action Potential and Ca2+ Transients in Ventricular Myocytes
To avoid potential effects of intracellular dialysis and Ca2+ buffering, membrane potentials and cytosolic Ca2+ transients were measured simultaneously in field-stimulated, fura-2AM-loaded ventricular myocytes using the perforated patch method (Figure 6A). At a pacing rate of 1 Hz, mean APD70 and APD90 were significantly shorter in I79N-Tg compared with control myocytes (Figure 6B, left). Mean resting potential (I79N-Tg −65±0.9 versus control −65±0.8 mV) and overshoot potential (I79N-Tg 21±3.1 versus control 27±2.7 mV) were not significantly different between control and I79N-Tg myocytes. At a more physiological pacing rate of 5 Hz (300 bpm), only APD70 remained significantly shorter in I79N-Tg compared with control myocytes. Overall action potential wave-form was reminiscent of those recorded with microelectrode in the intact heart of either group (compare Figure 4).
As illustrated in Figure 6A, Ca2+ transients of I79N-Tg myocytes were significantly smaller (fluorescent amplitude F340/F410, I79N 0.44±0.06, n=37 cells from 6 hearts, versus control 0.85±0.07, n=112 cells from 15 hearts [7 WT-Tg and 6 Non-Tg]; P<0.01), and their rate of decay (τ) was significantly slower compared with control myocytes (I79N-Tg 374±28 versus control 191±11 ms; P<0.001). In contrast, SR Ca2+ content (estimated from caffeine-induced Ca2+ transients) and diastolic [Ca2+] were not statistically different between I79N-Tg and control myocytes (compare also Figure 8). Increasing the pacing rate to 5 Hz increased the amplitude and accelerated the rate of decay in both groups, but mean values remained significantly different (fluorescent amplitude F340/F410, I79N 0.92±0.11 versus control 1.22±0.10; P<0.05; τ, I79N-Tg 204±22 versus control 102±5 ms; P<0.001). As a result of the slower Ca2+ transient decay kinetics, faster pacing increased diastolic [Ca2+] significantly more in I79N-Tg myocytes compared with control myocytes (percent increase, 27±2% versus 19±1%; P<0.01). At the same time, APD70 and APD90 remained unchanged in I79N-Tg myocytes, but significantly shortened in control myocytes (compare left and right panels of Figure 6B). Together with the results from highly buffered myocytes (Figure 5), these data suggest that changes in cytosolic [Ca2+] may contribute to the altered action potential morphology of I79N mice, possibly by causing differences in Na+-Ca2+ exchanger current.
To test this hypothesis directly, extracellular Na+ was rapidly replaced with Li+ (Figure 7A). This maneuver allows influx of Li+ through Na+ channels, but blocks Ca2+ extrusion on the exchanger.16 The block of inward exchanger current by Li+ abolished the differences in APD70 of transgenic and control mice (Figure 7B). Note that in the presence of Li+, the action potential is much shorter despite a much larger Ca2+ transient, suggesting that the inward exchanger current significantly contributes to repolarization of the action potential. In addition, Figure 7A (right) shows that enhanced Ca2+ transients produce longer action potentials after the washout of Li+. These data, together with those obtained during high Ca2+ buffering (Figure 5), suggest that changes in magnitude and kinetics of Ca2+ transients regulate the action potential of I79N transgenic mice via the Na+-Ca2+ exchanger.
Measurement of Ca2+ Transients in Response to Isoproterenol
Exposure to isoproterenol (500 nmol/L) increased peak Ca2+ transients more dramatically in I79N-Tg than in control myocytes (Figures 8A and 8B), such that there were no statistical differences between the two groups any longer (F340/F410, I79N-Tg 1.15±0.13 versus control 1.21±0.19; P=0.8; Figure 8C). As expected, isoproterenol also significantly accelerated the decay rate of Ca2+ transients in both groups (Figure 8D), but the decay rate of Ca2+ transients remained significantly slower in I79N-Tg versus control myocytes (τ, 111±10 versus 65±3 ms; P<0.01). Similar to the data obtained with fast pacing, isoproterenol significantly increased diastolic [Ca2+] in I79N-Tg versus control myocytes (Figure 8E). In addition, isoproterenol significantly increased the SR Ca2+ content of I79N-Tg versus control myocytes (Figure 8F). Thus, the data suggest that in the presence of isoproterenol, [Ca2+]i remains elevated for longer periods of time due to the slower rate of cytosolic Ca2+ decay in I79N-Tg compared with control myocytes.
Our data demonstrate for the first time that expression of mutant troponin T causes ventricular arrhythmias in freely moving mice (Figure 1), anesthetized mice (Figure 2), and in isolated hearts (Figure 3). These arrhythmias are triggered by stress or catecholamine stimulation and can be also evoked by increasing [Ca2+]o (Figure 3). Action potential remodeling of I79N-Tg hearts (Figure 4) and myocytes (Figure 6) is likely the result of changes in cytosolic Ca2+ transients (Figures 6 and 8⇑), which cause differential activation of Ca2+-dependent Na+-Ca2+ exchange (Figure 7). The combination of slow rate of cytosolic Ca2+ decay, especially when Ca2+ transients are greatly enhanced by isoproterenol (Figure 8), and reduction of IK1 (Figure 5D), may further contribute to the high frequency of stress-induced ventricular arrhythmias observed I79N mice.
Mechanism of Action Potential Remodeling
In highly Ca2+-buffered I79N-Tg and control cells, we found no significant differences in APD (Figure 5A), Ca2+ current (Figure 5B), or transient outward K+ current (Figure 5C), but IK1 was downregulated (Figure 5D). On the other hand, action potentials of poorly Ca2+-buffered I79N-Tg myocytes were significantly longer (Figure 6) and had an altered morphology (shorter APD70) similar to those measured in intact I79N-Tg hearts (Figure 4). Because the size of Ca2+ transients, peaking between APD50 and APD70, is significantly reduced in I79N-Tg myocytes (Figure 6A), it is likely that the Ca2+-extrusion mode of the exchanger would be suppressed, thereby resulting in shorter APD50 and APD70 in I79N-Tg myocytes. Conversely, because the decay of the Ca2+ transient is slow in I79N-Tg myocytes, [Ca2+]i is the same or higher in the terminal phase of the action potential (Figure 6A), resulting in unchanged APD90 at faster beating rates (Figures 4B and 6⇑B). Supporting this idea, block of Ca2+ efflux on the Na+-Ca2+ exchanger using Li+ replacement of [Na+]o eliminated the differences in APD70 (Figure 7). Together, these results suggested that [Ca2+]i strongly modulates the murine ventricular action potential via the exchanger, as also reported for rat ventricular myocytes.17
Mechanism of Ventricular Arrhythmias
There are several possibilities how the remodeling of repolarization wave-shape (Figure 4), induced by differences in Ca2+ transients (Figures 5 to 8⇑⇑⇑), could be arrhythmogenic. (1) Differences in repolarization wave-shape may contribute to larger spatial heterogeneity of refractoriness, which can cause reentrant ventricular tachacardia.18 (2) The slow decay of Ca2+ transient may lead to increased calmodulin kinase 2 activation, shown to cause early afterdepolarizations in mice,19 even in the absence of APD prolongation. (3) The combined decrease of IK1 (Figure 5) and increase of diastolic [Ca2+]i at fast pacing rates or in presence of isoproterenol (Figure 8) may trigger delayed afterdepolarizations,20 and/or spontaneous Ca2+ oscillations.21 (4) Abnormal Ca2+ handling may in itself contribute to the initiation of reentrant arrhythmias by mechanisms distinct from enhanced dispersion of refractoriness or triggered activity, as recently demonstrated in a murine heart failure model.22 Exactly which mechanism(s) are responsible for the ventricular tachycardia of I79N-Tg mice remains to be determined.
Mechanism of Altered Ca2+ Transients
The altered Ca2+ transients in I79N-Tg myocytes (Figures 6 and 8⇑) could be a direct consequence of the increased Ca2+ sensitivity of I79N-mutant myofilaments,11 because Ca2+ binding to myofilaments represents a substantial portion of cytoplasmic Ca2+ buffering.9 Experimentally increasing cytoplasmic Ca2+ buffering also results in reduced and prolonged Ca2+ transients.23 Likewise, transgenic mice expressing a TnI mutation with increased myofilament Ca2+ sensitivity also show depressed and prolonged Ca2+ transients.24 Based on experimental data from skinned fibers expressing the TnT-I79N mutation, mathematical modeling predicts smaller and slowly decaying Ca2+ transients in I79N compared with wild-type TnT fibers, assuming unchanged SR Ca2+ release and sarcolemmal Ca2+ flux.11 Thus, our finding of smaller and slower Ca2+ transients (Figures 6 and 8⇑) with no change in L-type Ca2+ current (Figure 5B) and SR Ca2+ content (Figure 8F) suggests that initially more Ca2+ is bound to TnT-I79N containing myofilaments, but later on, as the muscle starts to relax, the additional Ca2+ that comes off the myofilaments produces the slower decay of Ca2+ transients. It may seem somewhat surprising that peak caffeine-induced Ca2+ transients (Figure 8F) were not depressed in I79N myocytes. However, it should be noted that, unlike the fast twitch Ca2+ transients (Figure 8A), a sustained rise in [Ca2+]i in the presence of caffeine would almost fully saturate TnC Ca2+ binding sites (>90% assuming peak caffeine-induced free [Ca2+]i of 1 μmol/L9 and high cooperativity of Ca2+ binding and myofilament activation in intact muscle25). Under such conditions, the TnT-I79N mutation may change the total cytosolic Ca2+ buffering capacity by less than 5% (given that TnT-I79N left-shifts pCa50 by 0.211). Alternatively, the data could simply imply that I79N-Tg cells have decreased SR Ca2+ release and depressed cytoplasmic Ca2+ removal. Thus, future studies will need to independently quantify SR Ca2+ content and Ca2+ buffering capacity of I79N-Tg myocytes and address possible contribution of Na+-Ca2+ exchanger and SR Ca2+-ATPase to the altered Ca2+ homeostasis described here.
Limitations and Potential Implications
Murine and human cardiac electrophysiology differ substantially,26 and action potential changes in response to expression of mutant TnT may be different in humans. Furthermore, ventricular tachyarrhythmias were nonsustained in I79N-Tg mice and did not cause significant mortality. As previously reported,10 I79N-Tg mice died after isoproterenol injection with complete heart block captured on ECG records. Although heart block did not occur in freely moving mice or in isolated perfused hearts, and isoproterenol-induced cardiac contractile dysfunction preceded the heart block,10 we cannot exclude that atrioventricular conduction abnormalities may also contribute to sudden deaths related to the I79N mutation. Nevertheless, our data provide the first direct evidence that the TnT-I79N mutation could cause ventricular tachycardia in an in vivo mouse model, even in absence of significant cardiac hypertrophy or fibrosis.
This work was supported in part by AHA SDG 0130285N (B.C.K.), PhRMA Foundation grant (B.C.K.), NIH grants HL42325 and HL674154 (J.D.P.) and HL16152 (M.M.), Collaborative Research Center 556 of the German Research Council, project Z2 (P.K., L.F.), and IZKF University of Münster, project ZPG4 (P.K., L.F.).
↵*Both authors contributed equally to this work.
Original received October 8, 2002; resubmission received December 16, 2002; revised resubmission received January 23, 2003; accepted January 23, 2003.
Varnava AM, Elliott PM, Baboonian C, Davison F, Davies MJ, McKenna WJ. Hypertrophic cardiomyopathy: histopathological features of sudden death in cardiac troponin T disease. Circulation. 2001; 104: 1380–1384.
Maron BJ, Shen WK, Link MS, Epstein AE, Almquist AK, Daubert JP, Bardy GH, Favale S, Rea RF, Boriani G, Estes NA, 3rd, Spirito P. Efficacy of implantable cardioverter-defibrillators for the prevention of sudden death in patients with hypertrophic cardiomyopathy. N Engl J Med. 2000; 342: 365–373.
Bers DM. Calcium Sources and Sinks. In: Excitation and Contraction Coupling and Cardiac Contractile Force, 2nd ed. Dordrecht/Boston/London: Kluwer Academic Publishers; 2001: 39–56.
Knollmann BC, Blatt SA, Horton K, de Freitas F, Miller T, Bell M, Housmans PR, Weissman NJ, Morad M, Potter JD. Inotropic stimulation induces cardiac dysfunction in transgenic mice expressing a troponin T (I79N) mutation linked to familial hypertrophic cardiomyopathy. J Biol Chem. 2001; 276: 10039–1048.
Miller T, Szczesna D, Housmans PR, Zhao J, deFreitas F, Gomes AV, Culbreath L, McCue J, Wang Y, Xu Y, Kerrick WG, Potter JD. Abnormal contractile function in transgenic mice expressing an FHC-linked Troponin T (179N) mutation. J Biol Chem. 2001; 276: 3743–3755.
Boersma L, Zetelaki Z, Brugada J, Allessie M. Polymorphic reentrant ventricular tachycardia in the isolated rabbit heart studied by high-density mapping. Circulation. 2002; 105: 3053–3061.
Wu Y, Temple J, Zhang R, Dzhura I, Zhang W, Trimble R, Roden DM, Passier R, Olson EN, Colbran RJ, Anderson ME. Calmodulin kinase II and arrhythmias in a mouse model of cardiac hypertrophy. Circulation. 2002; 106: 1288–1293.
Bers DM. Calcium and cardiac rhythms: physiological and pathophysiological. Circ Res. 2002; 90: 14–17.
Lakatta EG. Functional implications of spontaneous sarcoplasmic reticulum Ca2+ release in the heart. Cardiovasc Res. 1992; 26: 193–214.
London B, Baker LC, Lee JS, Shusterman V, Choi BR, Kubota T, McTiernan CF, Feldman AM, Salama G. Calcium-dependent arrhythmias in transgenic mice with heart failure. Am J Physiol Heart Circ Physiol. 2003; 284: H431–H441.
MacGowan GA, Du C, Cowan DB, Stamm C, McGowan FX, Solaro RJ, Koretsky AP, Del Nido PJ. Ischemic dysfunction in transgenic mice expressing troponin I lacking protein kinase C phosphorylation sites. Am J Physiol Heart Circ Physiol. 2001; 280: H835–H843.
Gao WD, Backx PH, Azan-Backx M, Marban E. Myofilament Ca2+ sensitivity in intact versus skinned rat ventricular muscle. Circ Res. 1994; 74: 408–415.