In cardiac-specific Na+-Ca2+ exchanger (NCX) knockout (KO) mice, the ventricular action potential (AP) is shortened. The shortening of the AP, as well as a decrease of the L-type Ca2+ current (ICa), provides a critical mechanism for the maintenance of Ca2+ homeostasis and contractility in the absence of NCX (Pott C, Philipson KD, Goldhaber JI. Excitation-contraction coupling in Na+-Ca2+ exchanger knockout mice: reduced transsarcolemmal Ca2+ flux. Circ Res 97: 1288–1295, 2005). To investigate the mechanism that underlies the accelerated AP repolarization, we recorded the transient outward current (Ito) in patch-clamped myocytes isolated from wild-type (WT) and NCX KO mice. Peak Ito was increased by 78% and decay kinetics were slowed in KO vs. WT. Consistent with increased Ito, ECGs from KO mice exhibited shortened QT intervals. Expression of the Ito-generating K+ channel subunit Kv4.2 and the K+ channel interacting protein was increased in KO. We used a computer model of the murine AP (Bondarenko VE, Szigeti GP, Bett GC, Kim SJ, and Rasmusson RL. Computer model of action potential of mouse ventricular myocytes. Am J Physiol Heart Circ Physiol 287: 1378–1403, 2004) to determine the relative contributions of increased Ito, reduced ICa, and reduced NCX current (INCX) on the shape and kinetics of the AP. Reduction of ICa and elimination of INCX had relatively small effects on the duration of the AP in the computer model. In contrast, AP repolarization was substantially accelerated when Ito was increased in the computer model. Thus, the increase in Ito, and not the reduction of ICa or INCX, is likely to be the major mechanism of AP shortening in KO myocytes. The upregulation of Ito may comprise an important regulatory mechanism to limit Ca2+ influx via a reduction of AP duration, thus preventing Ca2+ overload in situations of reduced myocyte Ca2+ extrusion capacity.
- genetically altered mice
- cardiac myocytes
- short QT interval
- transient outward current
in cardiac myocytes, the contractile cycle is initiated by Ca2+ entry via L-type Ca2+ channels. To maintain Ca2+ balance, the Na+-Ca2+ exchanger (NCX) completes the cycle by extruding the same amount of Ca2+ that entered through the Ca2+ channels (1, 10, 24). NCX is therefore viewed as an essential component of cardiac excitation-contraction coupling (ECC), and functional myocardium in the absence of NCX would seem impossible. Nevertheless, cardiac-specific NCX knockout (KO) mice are viable to adulthood with almost normal cardiac performance (13, 26).
We reported previously (27) that there are two fundamental mechanisms of adaptation to the absence of NCX: 1) a decrease in L-type Ca2+ current (ICa) with an increased ECC gain and 2) a shortened and more rapidly repolarizing action potential (AP) that further reduces Ca2+ influx. The more rapid repolarization of the AP renders it as effective as the wild-type (WT) AP in triggering sarcoplasmic reticular (SR) Ca2+ release. These two mechanisms independently but synergistically limit Ca2+ influx so that Ca2+ homeostasis can be maintained by a low-capacity efflux mechanism (i.e., the plasma membrane Ca2+ ATPase) while at the same time maintaining contractility.
It is important to understand the mechanism responsible for the shortened AP in KO mice, since this mechanism plays a key role in assuring both Ca2+ homeostasis and effective ECC in the absence of NCX. Both the elimination of NCX current (INCX) and the reduction of ICa could potentially explain the abbreviated AP in KO myocytes. However, the dominant repolarizing current in cardiac ventricular myocytes is the transient outward current (Ito) (20). Recent studies (8, 23, 35–37) have suggested that some of the cardiac K+ channel subunits that generate Ito are tightly regulated by cytosolic Ca2+ in their expression and activity, thus making regulation of the AP by Ito in response to changes in Ca2+ produced by knockout of NCX an attractive possibility.
We report here that Ito is increased in NCX KO mice and that the expression of the Ito generating voltage-dependent K+ channel subunit Kv4.2 and the K+ channel interacting protein (KChIP) are upregulated. With the use of computer modeling, we further demonstrate that Ito upregulation is the main determinant of AP shortening in KO mice. We hypothesize that altered Ca2+ handling in NCX KO mice leads to upregulation of Ito via an unknown mechanism. This in turn reduces AP duration and limits Ca2+ influx. Ito upregulation may comprise an important negative feedback mechanism against Ca2+ overload in situations of reduced myocyte Ca2+ extrusion capacity.
Generation of transgenic mice.
NCX cardiac-specific KO mice were generated using Cre/Lox technology, as previously described (13). NCX KO mice used for this study were between 8 and 12 wk of age and did not display any gross pathology or signs of heart failure. No increase in sudden cardiac death was observed in NCX KO mice vs. WT mice.
ECG data were collected in awake, freely moving mice by implanting radio telemetry devices (model TA10ETA-F20, Transoma-Data Science International) as previously described (34). Transmitter units were implanted in the peritoneal cavity of anesthetized mice, and the two electrical leads were secured near the apex of the heart and the right acromion in a lead II orientation. Unfiltered ECG data for this study were collected for 30 s every hour for at least 1 mo. RR and QT intervals were calculated from all of the ECG waveforms with the Transoma-Data Science International analysis packages (ART 3.1 and Physiostat 4.01).
Isolation of ventricular myocytes from adult mouse hearts.
Myocytes were isolated from WT and KO animals with the use of the collagenase/protease digestion method reported previously (27) and in accordance with the guidelines of the UCLA Office for Protection of Research Subjects. The atrium and base of the ventricle were discarded. Isolated cardiomyocytes were stored for up to 6 h at room temperature in modified Tyrode solution containing (in mM) 136 NaCl, 5.4 KCl, 10 HEPES, 1.0 MgCl2, 0.33 NaH2PO4, 1.0 CaCl2, 10 glucose, pH 7.4, with NaOH.
To record whole cell membrane currents, we placed the cells in an experimental chamber (0.5 ml) mounted on the stage of a Nikon Diaphot inverted microscope. A heated bath solution (26°C) continuously perfused the chamber. Patch electrodes were pulled from borosilicate glass (model TW150F-3; World Precision Instruments, Sarasota, FL) on a horizontal puller (model P-97; Sutter Instruments, Novato, CA). The fire-polished electrodes had a tip diameter of 2–3 μm and a resistance of 1–2 MΩ. The pipette solution contained (in mM) 130 KCl, 1 MgCl, 5.4 NaCl, 10 HEPES, 5 MgATP, 0.05 cAMP, 2.8 phosphocreatine, pH 7.2, with KOH with modifications described below. We recorded Ito as the whole cell membrane current using an Axopatch 200 or 200B patch-clamp amplifier (Molecular Devices, Sunnyvale, CA) in voltage-clamp mode, and a Digidata 1322A (Molecular Devices) data-acquisition system controlled by pCLAMP 9 software (Molecular Devices). We applied series resistance compensation to all recordings.
Determination of cellular phenotype.
Ten to twenty percent of the myocytes isolated from KO hearts have a WT phenotype (13). To identify and exclude such cells from this study, myocytes used from KO mice were tested for INCX either during rapid application of caffeine (27) or during SR Ca2+ release induced by a brief depolarization (28). Those cells with an inward exchange current were excluded from the KO group.
Western blot analysis.
The particulate fraction of mouse ventricles (60–100 μg) was dissolved in SDS-reducing buffer. Proteins were separated on 8–10% SDS-PAGE and transferred to nitrocellulose. Blots were probed with antibodies specific for the K+ channel subunits Kv4.3, Kv4.2, KChIP, Kv2.1, Kv1.5, and Kv1.4 (Alomone Laboratories, Jerusalem, Israel). Additional antibodies against Kv2.1, Kv1.5, and Kv1.4 were provided by the Neuromab facility at the University of California Davis. These antibodies have been used successfully in previous studies (12, 37) to detect the corresponding K+ channel subunits in mouse myocardium. Blots were developed using chemiluminescence (MEN Life Science, Hercules, CA) and quantitated by laser densitometry.
To model effects of INCX, ICa, and Ito on the AP, we used a computer model of the murine AP and its underlying ionic currents (4). To our knowledge, this is the only published computer model specifically dealing with the murine AP. The model was downloaded from the CellML web site (www.cellml.org) and was run on a Pentium-based personal computer. Changes in peak current and inactivation kinetics were modeled in the voltage-clamp mode for INCX, ICa, and Ito until we achieved similarity with changes observed in vitro in this and previous studies (13, 27, 28). The modified currents were then introduced into the model's AP pacing mode, and data were obtained after AP kinetics reached steady state during 1-Hz pacing. The original model has two different parameter regimes that model myocytes from two different regions of the murine heart, the apex, and the septum (4). In this study, we report results based on simulated APs using the apex mode of the model, but similar results were also observed when modeling was conducted in the septum mode.
Data are expressed as means ± SE. Student's unpaired t-test was used for direct comparisons of WT vs. KO. When Ito was tested in a range of voltages in each group, we used two-way ANOVA with the Tukey-Fisher least-significant difference post hoc test (JMP 5.01a; SAS Institute, Cary, NC).
Short QT interval in KO mice.
Figure 1A shows representative ECG tracings from WT and KO mice. The duration of the QT interval was significantly shortened in KO mice compared with their WT littermates during daytime recordings (WT, 51 ± 3 ms; n = 8; KO, 39 ± 1 ms; P < 0.05) (see Fig. 1B). There was no significant difference in the RR interval (Fig. 1B). Similar results were observed during nighttime measurements. The abbreviated QT interval is consistent with the reduced AP duration in single cells that we have described previously (13, 27) and confirms the effects of early repolarization at the tissue level.
Increased Ito in NCX KO mice.
Ito was recorded using the whole cell patch-clamp technique during depolarizations from −80 mV to a family of test potentials between −60 and +50 mV. The bath solution contained (in mM) 136 NaCl, 5.4 KCl, 10 HEPES, 1.0 MgCl2, 0.33 NaH2PO4, 1.0 CaCl2, 10 glucose, pH 7.4 with NaOH. Tetrodotoxin (10 μM) was added to block INa, and nifedipine (0.2–2 μM) was added to block ICa and Ca2+-gated Cl− currents (35). Figure 2A shows the stimulation protocol (top panel) and representative traces for Ito in KO and WT myocytes (middle panel). In KO myocytes, Ito amplitude was increased over the full range of test potentials. At +40 mV, the peak Ito amplitude of KO myocytes was 183% of that of WT myocytes (KO, 53.0 ± 3.7 pA/pF; WT, 28.9 ± 3.6 pA/pF; P < 0.01 by two-way ANOVA; Fig. 2B).
Ito consists of fast and slow components (6, 20) that can be dissected by exponential analysis (34). Double exponential curves were fitted to the Ito tracings. In KO myocytes, the amplitude of the slow component of Ito was increased (Fig. 2C, top panel), but the increase in the fast component did not reach statistical significance (P = 0.36). The inactivation rates of both the fast and slow components were slowed in KO (Fig. 2C, bottom panel).
Because Ca2+ handling is drastically altered in KO mice (26), we investigated the possibility that differences in cytosolic Ca2+ concentration or kinetics are responsible for the observed increase of Ito in KO. This might occur, for example, via activation of Ca2+-gated Cl− or K+ channels (35, 36). We therefore repeated the measurements of Ito after dialyzing the myocytes with the fast Ca2+ chelator BAPTA (10 mM) by including it in the patch pipette (5, 27, 30). Buffering cytosolic Ca2+ with BAPTA did not prevent the observed increase of Ito in KO myocytes (peak Ito amplitude at +40 mV for KO was 200 ± 35% of that for WT; P < 0.01; Fig. 2A, bottom panel). These data show that Ito is upregulated in myocytes from NCX KO mice and that this upregulation is not directly dependent on cytosolic Ca2+.
Expression of Ito generating K+ channel subunits in NCX KO hearts.
Ito is generated by several different K+ channels composed of multiple subunits (21). To further investigate the mechanism that underlies the increase of Ito in NCX KO mice, we performed immunoblots in cardiac membrane fractions from WT and KO mice using specific antibodies against the K+ channel subunits Kv4.3, Kv4.2, KChIP, Kv2.1, Kv1.5, and Kv1.4 (37). Figure 3 shows representative immunoblots of WT and KO homogenates (bottom panel) and the corresponding densitometric quantification (top panel). In KO mice, expression of both Kv4.2 (169 ± 15% vs. WT; P < 0.05) and KChIP (140 ± 16% vs. WT; P < 0.05) was significantly increased compared with WT mice. No significant changes were observed for the expression levels of Kv4.3, Kv2.1, Kv1.5, and Kv1.4.
The upregulation of Ito provides an explanation for the abbreviated AP in NCX KO mice. However, both INCX and ICa are reduced in NCX KO mice (13, 27, 28), which could also contribute to the reduced AP duration. To determine the relative contributions of altered INCX, ICa, and Ito to the accelerated AP repolarization we used a validated computer model of the murine action potential (4). Figure 4 shows the steady-state simulation of a WT AP (solid line) and the effects of different modifications (dashed line).
First, we eliminated INCX from the model (Fig. 4A). This had very minor effects on AP repolarization seen only toward the end of the AP. We next reduced ICa amplitude by 50% and accelerated the ICa inactivation rate using values from our previously published observations in KO myocytes (27, 28) (Fig. 4B). Again, this only modestly accelerated AP repolarization. In contrast, AP repolarization was substantially accelerated when we increased peak Ito amplitude in the model by 78% and slowed inactivation kinetics in the model as observed in vitro in this study (Fig. 4C). When all three alterations were introduced into the model (Fig. 4D), the resulting AP was similar to that produced by simply upregulating Ito in the model (Fig. 4C). The effects of these manipulations on the duration to 50% (APD50) and 90% (APD90) repolarization of the AP are summarized in Table 1. These data suggest that the observed shortening of the KO AP in vitro (13, 27) is caused mainly by the increase in Ito with only modest contributions by the reduction of INCX and ICa.
The ventricular AP is shortened in NCX KO mice (13, 27). This appears to be a crucial mechanism to adapt ECC to the absence of NCX. The abbreviated AP limits Ca2+ influx into the myocyte so that Ca2+ homeostasis can be maintained by low-capacity extrusion mechanisms such as the sarcolemmal Ca2+ ATPase (27). Remarkably, the shortened KO AP is as effective in triggering SR release as the longer WT AP, presumably because the more rapid repolarization increases the efficiency of cardiac ECC by mechanisms described in detail previously (29). Below, we discuss the possible mechanisms that underlie the shortened AP in KO myocytes and speculate on the causal pathway that might lead from knockout of NCX to the observed accelerated AP kinetics.
Increased Ito in myocytes from NCX KO mice.
Ito is an important repolarizing current in cardiac cells. It is activated early in repolarization in multiple species but is particularly robust in rats and mice (21). We have demonstrated in this study that Ito is increased significantly in ventricular NCX KO myocytes. Previous studies have demonstrated that cytosolic Ca2+ can regulate the activity of some Ito-generating channels, such as Ca2+-gated K+ or Cl− channels (35, 36) or the accessory subunit KChIP (8). Ca2+ handling is changed in the absence of NCX (27, 28) and thus an altered activity of Ca2+-gated outward currents is an attractive explanation for the observed increase in Ito. However, heavy buffering of cytosolic Ca2+ with BAPTA, as used previously (5, 28, 30), did not affect the characteristic increase of Ito in KO myocytes (Fig. 2). These results suggest that increased Ito in KO mice is not an acute consequence of Ca2+ modulation of Ito-generating channels.
Chronic alterations of cytosolic Ca2+ can regulate the expression of the Ito-generating K+ channel subunits Kv4.2 and KChIP (23, 37), and an involvement of calcineurin in this process has been suggested (23). Indeed, we observe increased protein expression of these subunits (Fig. 3), which is consistent with the increased amplitude and the delayed inactivation kinetics of Ito observed in NCX KO myocytes. However, exponential analysis of our data indicates a significant increase in only the slow component of Ito (Fig. 2C), whereas Kv4.2 reportedly contributes mostly to the fast component (11, 18, 21). Thus we do not yet have mechanistic insight into the upregulation of the slow component of Ito induced by knockout of the NCX.
A variety of studies using transgenic mice have demonstrated a reduced or eliminated expression of K+ channel subunits that led to a decrease of Ito and a prolongation of the AP (11, 12, 16–18, 37). To our knowledge, NCX KO mice are the only animal model with a genetic modification that exhibits an increased expression of cardiac K+ channel subunits, an increase in Ito, and an accelerated AP repolarization.
The upregulated expression of Kv4.2 and KChIP is likely due to altered Ca2+ handling as demonstrated in other models (23, 37). The alterations in Ca2+ handling observed in the NCX KO model are not manifest as changes of global resting or systolic Ca2+ but primarily in a reduction of transsarcolemmal Ca2+ fluxes (27) and possibly subsarcolemmal Ca2+ concentration (28). These conditions are difficult to monitor or modulate under experimental conditions. Nevertheless, the upregulation of Ito and the subsequent reduction of Ca2+ influx in response to knockout of NCX provide further evidence of an interrelationship between cardiac Ca2+ handling and the repolarizing Ito system and for the plasticity of cardiac ECC in general.
Effects of elimination of INCX, reduced ICa, and increased Ito on AP repolarization.
Inward NCX current can contribute to the plateau phase of the AP (22, 32, 38), and thus one might expect the elimination of INCX in KO myocytes to account for the faster AP repolarization. Similarly, ICa is an important component of the AP plateau. In KO mice ICa is reduced, possibly because the absence of NCX activity leads to an increase in subsarcolemmal Ca2+, which reduces ICa through Ca2+-dependent inactivation (28). This reduction of ICa would seem to be an explanation for the abbreviated KO AP (14, 33). However, accelerated repolarization of the KO AP is evident even in the early phases of the AP (13, 27). Thus, it is questionable whether elimination of INCX and/or reduced ICa substantially contributes to the accelerated AP repolarization in KO mice. Consistent with this concern, eliminating INCX (Fig. 4A) or reducing ICa (Fig. 4B) in a computer model of the murine AP had only very small effects on AP repolarization and AP duration (Table 1).
The repolarizing current that is active during the initial phases of AP repolarization is Ito. Especially in mouse myocytes, Ito dominates both the early and late phases of the AP (6, 21). Accordingly, we see a strong acceleration of AP repolarization when modeling the increase of Ito that we observe in NCX KO myocytes (Fig. 4C). We thus conclude that the abbreviation of the AP in NCX KO mouse myocytes is primarily mediated by the increase in Ito and not by the reduction of INCX or ICa.
Species-dependent differences in AP repolarization.
In mouse cardiomyocytes, AP repolarization is extremely rapid with only a brief plateau phase at very negative potentials (6, 7, 35). The plateau-maintaining currents INCX and ICa contribute only mildly to AP repolarization in computer simulations (Fig. 4, A and B), whereas there is a high sensitivity to the magnitude of Ito (Fig. 4C). This stands in marked contrast to other species, such as the human, rabbit, or guinea pig, that have a much more pronounced plateau phase of the AP and a smaller dependence on Ito. Manipulations of NCX and ICa are of greater consequence for AP repolarization in these species (9, 21, 32).
Physiological and pathophysiological implications.
We hypothesize that NCX KO mice upregulate Ito in response to the severely reduced Ca2+ extrusion capacity of this model (27). The increased Ito shortens the AP, which both reduces the QT interval on ECG and limits Ca2+ influx into the myocyte by almost 60% (27). The reduction in Ca2+ influx helps KO mice survive in the absence of NCX. Thus upregulation of Ito may comprise a protective mechanism against Ca2+ overload in situations of reduced Ca2+ extrusion capacity.
In several models of heart failure, NCX expression is increased (25) whereas SR-Ca2+ ATPase activity is reduced (3, 31). This creates a situation in which the competition for Ca2+ removal from the cytosol is shifted toward NCX. A further characteristic of failing myocardium is a downregulation of Ito and a prolongation of the AP (2, 15, 19, 20). It is unknown whether the upregulation of NCX and the reduced SR-Ca2+ ATPase and Ito activity are causal or consequential to heart failure. Interestingly, NCX KO mice mirror these changes by exhibiting increased Ito, a shortened AP, and preserved SR Ca2+ stores. Our findings suggest that alteration of NCX expression can influence Ito, and a similar mechanism could potentially contribute to modulation of Ito in other situations.
This research was supported by Köln Fortune and the German Research Foundation Grants DFG PO 1004\1-1 and PO 1004/1-2 (to C. Pott), National Institutes of Health Grants HL-70828 (to J. I. Goldhaber) and HL-48509 (to K. D. Philipson), and the Laubisch Foundation.
We appreciate helpful discussions with Dr. L. Xie.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society