Cellular Signaling Underlying Atrial Tachycardia Remodeling of L-type Calcium Current
Atrial tachycardia (AT) downregulates L-type Ca2+ current (ICaL) and causes atrial fibrillation–promoting electric remodeling. This study assessed potential underlying signal transduction. Cultured adult canine atrial cardiomyocytes were paced at 0, 1, or 3 Hz (P0, P1, P3) for up to 24 hours. Cellular tachypacing (P3) mimicked effects of in vivo AT: decreased ICaL and transient outward current (Ito), unchanged ICaT, IKr, and IKs, and reduced action potential duration (APD). ICaL was unchanged in P3 at 2 and 8 hours but decreased by 55±6% at 24 hours. Tachypacing caused Ca2+i accumulation in P3 cells at 2 to 8 hours, but, by 24 hours, Ca2+i returned to baseline. Cav1.2 mRNA expression was not altered at 2 hours but decreased significantly at 8 and 24 hours (32±4% and 48±4%, respectively) and protein expression was decreased (47±8%) at 24 hours only. Suppressing Ca2+i increases during tachypacing with the ICaL blocker nimodipine or the Ca2+ chelator BAPTA-AM prevented ICaL downregulation. Calcineurin activity increased in P3 at 2 and 8 hours, respectively, returning to baseline at 24 hours. Nuclear factor of activated T cells (NFAT) nuclear translocation was enhanced in P3 cells. Ca2+-dependent signaling was probed with inhibitors of Ca2+/calmodulin (W-7), calcineurin (FK-506), and NFAT (INCA6): each prevented ICaL downregulation. Significant APD reductions (≈30%) at 24 hours in P3 cells were prevented by nimodipine, BAPTA-AM, W-7, or FK-506. Thus, rapid atrial cardiomyocyte activation causes Ca2+ loading, which activates the Ca2+-dependent calmodulin–calcineurin–NFAT system to cause transcriptional downregulation of ICaL, restoring Ca2+i to normal at the cost of APD reduction. These studies elucidate for the first time the molecular feedback mechanisms underlying arrhythmogenic AT remodeling.
Atrial fibrillation (AF) is the most common clinical tachyarrhythmia, with an incidence that increases with age and a significant association with cardiovascular morbidity and mortality.1 AF causes electrophysiological changes, primarily resulting from the rapid atrial activation rates, that promote arrhythmia perpetuation.2,3 Atrial action potential (AP) duration (APD) shortening is a major contributor to refractoriness abbreviation, which is, in turn, a primary factor in AF promotion.4,5 Loss of APD and impaired APD rate adaptation are largely attributable to the associated ICaL reduction.4,6
Atrial remodeling is believed to have important therapeutic implications, and there is interest in developing antiremodeling therapies,7 but this approach is limited by an insufficient understanding of underlying mechanisms to allow for the definition of molecular targets. There is indirect evidence for a role of Ca2+ overload in the remodeling caused by atrial tachycardia (AT),8,9 and transcriptional downregulation of the Cav1.2 ICaL α-subunit contributes to ICaL downregulation.5,10 However, the signaling mechanisms coupling Ca2+ loading to ICaL downregulation are poorly understood. In this study, we used an in vitro model of paced canine atrial cardiomyocytes to assess (1) whether it mimics in vivo AT-induced cellular electrophysiological remodeling; and (2) if so, what intracellular signaling mechanisms are involved.
Materials and Methods
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
Single canine left atrial cells were isolated with previously described methods.11 Cardiomyocytes were kept in medium 199 and concentrated by centrifugation at 500 rpm (1 minute), and cell pellets were removed for culture.
Cells were plated at low density (≈104 cells/cm2) onto laminin-coated (20 μg/mL) glass coverslips and maintained at 37°C. After 4 hours, dead and unattached myocytes were removed and fresh medium was added.12 Pacing was accomplished with square wave, 5-ms pulses. Parallel 24-hour culture studies were performed with cells subjected to 1-Hz (P1) and 3-Hz (P3) pacing and no pacing (P0). For time course studies, results with 1- and 3-Hz pacing were compared to each other at each time point and to prepacing baseline. After cell culture with or without pacing, cells were transferred to superfusion baths for ionic current or AP recording or harvested and frozen for subsequent biochemical study. In some experiments, 1 μmol/L nimodipine, 10 μmol/L BAPTA-AM, 1 μmol/L W-7, 5 μmol/L FK-506, or 10 μmol/L INCA-6 was added to the culture medium and thoroughly washed out before recording currents or [Ca2+]i. Cell viability was measured by Trypan blue dye exclusion.
Ionic Current and AP Recording
All in vitro recordings were obtained at 37°C. The whole-cell perforated-patch technique was used to record APs and tight seal patch clamp to record ionic currents. Junction potentials averaged 15.9 mV and were corrected for APs only. For solution contents for recording of specific ionic currents and APs, see the expanded Materials and Methods section in the online data supplement.
Atrial cardiomyocytes were incubated with indo-1 AM (5 μmol/L) in 100 μmol/L pluronic F-127 and 0.5% DMSO for 3 to 5 minutes and then superfused with Tyrode’s solution. Ultraviolet light passing through a 340-nm interference filter was reflected into a ×40 oil immersion fluor objective for excitation of intracellular indo-1. Exposure of the cell to UV light was controlled with an electronic shutter to minimize photobleaching. Emitted light was reflected into a spectral separator, passed through parallel filters at 400 and 500 nm (±10 nm), detected by matched photomultiplier tubes, and electronically filtered at 60 Hz. Chamber background fluorescence was removed by adjusting the 400- and 500-nm channels to 0 over an empty field of view near the cell. Fluorescence signal ratios were digitized and converted to [Ca]i as previously described.13 Cells were paced with 10-ms 1.5-times threshold-voltage pulses.13
Calcineurin Activity and NFAT Imaging
Atrial cardiomyocyte samples were collected prepacing and after 2, 8, or 24 hours of 1- or 3-Hz pacing. Calcineurin phosphatase activity was measured with a commercial assay kit (Calbiochem). Atrial cardiomyocyte samples were homogenized in lysis buffer. Calcineurin activity was measured in phosphatase standard buffer and determined as the dephosphorylation rate of the R-II peptide. Free phosphate released from R-II peptide was measured at 620 nm.
To measure nuclear and cytoplasmic NFAT (nuclear factor of activated T cells) immunofluorescence, cells were fixed with 2% paraformaldehyde in PBS, blocked, and permeabilized with PBS containing 2% normal donkey serum, 2% BSA, and 0.1% Triton X-100. Cells were then incubated overnight at 4°C with primary antibodies (monoclonal mouse anti-NFATc3 and polyclonal rabbit anti-NFATc4, 1/200 in PBS with 1% normal donkey serum, 1% BSA, and 0.05% Triton-X100). Donkey anti-mouse Alexa fluor (emission peak, 555-nm) and donkey anti-rabbit Alexa fluor (488 nm) were corresponding secondary antibodies. Cells were then exposed to RNaseA (100 μg/mL) for 25 minutes at 37°C, followed by 3 washes and then incubated with ToPro3 for nuclear contour definition (1 μmol/L) for 45 minutes. Images were obtained with a Zeiss LSM-510 confocal microscopy system and deconvolved with point-spread functions acquired with the same parameters as the images of interest. Nuclear and cytosolic NFATc3 and NFATc4 staining was quantified from the total number of pixels within the corresponding region normalized to area. Measurements were obtained in 5 Z-stacks covering the maximum nuclear area to calculate mean densities for each cell.
Quantitative Real-Time PCR
Total RNA was isolated from 100- to 300-mg atrial cardiomyocyte samples with TRIzol, followed by chloroform extraction and isopropanol precipitation, DNase treatment, and quality control with polyacrylamide gel electrophoresis.14 First-strand cDNA was synthesized from 2 μg of total RNA with High Capacity cDNA Archive Kits. Real-time quantitative PCR (QPCR) was performed with 6-carboxy-fluorescein–labeled fluorogenic Cav1.2 TaqMan primers and probes and TaqMan universal master mix. Fluorescence signals were detected in duplicate, normalized to 18S ribosomal RNA, and quantified with MxPro QPCR software.
Western Blot Analysis
Paced atrial cardiomyocyte samples were homogenized in radioimmunoprecipitation assay buffer as previously described.15 The homogenate was centrifuged (15 000 rpm, 20 minutes, 4°C). The supernatant was used for protein concentration measurement by Bradford assay with BSA as a standard. For α1c subunit assessment, 40-μg protein samples were separated by 8% Na-dodecylsulfate polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and incubated with rabbit anti–cardiac Cav1.2, 1:100 and mouse antiglyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:20 000), followed by goat anti-mouse (1:2000) or goat anti-rabbit (1:5000) horseradish peroxidase–conjugated secondary antibodies. Signals were visualized with Western Lightning Chemiluminescence Reagent Plus and quantified by videodensitometry.
Clampfit 9.2, GraphPad Prism 4.0, IgorPro5.04B, MxPro-Mx 3000P, SPSS, and Origin 5.0 were used for data analysis. All data are expressed as means±SEM. Multiple group statistical comparisons were obtained by 2-way ANOVA, and individual group mean differences were evaluated by Student’s t tests with Bonferroni correction. A 2-tailed P<0.05 was considered statistically significant.
Cell capacitance averaged 142±7, 144±8, and 148±8 pF (n=35 per group; P=NS) in P0, P1, and P3 cells, respectively. Cell viability averaged ≈50% in P0 and P1 cells and was reduced to ≈30% in P3 cells (supplemental Table I). Cell dimensions (supplemental Table I) and morphologies (online data supplement, Figure I) did not change with tachypacing. After 24 hours of pacing at 0, 1, or 3 Hz, APs were recorded at frequencies between 0.1 and 2 Hz, allowing 1 minute at each frequency to reach steady state. Resting membrane potential and AP amplitude were not altered by tachypacing. Resting membrane potential averaged −72.5±2.1 mV (n=14 cells) in P0 cells compared with −74.1±1.5 mV (n=16) and −74.1±0.9 mV (n=19) in P1 and P3 cells, respectively (P=NS). AP amplitude at 1 Hz averaged 114±2 mV (n=14 cells) in P0 cells, 112±2 mV (n=16) in P1 cells, and 108±3 mV (n=19) in P3 cells (P=NS). Figure 1A shows representative APs recorded at 1 Hz from P0, P1, and P3 cardiomyocytes. APD90 (APD to 90% repolarization) was significantly reduced and APD rate adaptation blunted in P3 cells (Figure 1B). APDs were not significantly different in P0 versus P1 cells.
Ionic Current Changes
Figure 2A shows ICaL recordings on 200-ms depolarizing steps from −50 to +10 mV. ICaL density was significantly decreased by 24-hour tachypacing at test potentials between −10 mV and +40 mV (Figure 2B). For example, ICaL at +10 mV averaged −5.0±0.5 pA/pF (n=13), −4.8±0.6 pA/pF (n=16), and −2.0±0.1 pA/pF (n=13) in P0, P1, and P3 cells, respectively. ICaL inactivation voltage dependence was determined with 1000-ms prepulses to voltages between −90 and +50 mV, followed by 300-ms test pulses to +10 mV. Activation voltage dependence was obtained from data obtained as illustrated in Figure 2B, based on the relation Iv=Imax(V−Vr)(Gv/Gmax), where Iv and Gv are current and conductance at voltage V; Imax and Gmax are maximum current and conductance, and Vr is the reversal potential. Vr was determined from the horizontal axis intercept of the ascending limb of the ICaL–voltage relation. The voltage dependencies of ICaL activation and inactivation were unaffected (Boltzmann fits) in tachypaced cells (Figure 2C). Voltage (V1/2) for half-maximum activation averaged −3.1±0.7 (n=13), −3.5±0.9 (n=16), and −3.6±0.8 mV (n=13) in P0, P1, and P3 cells, respectively (P=NS). The inactivation V1/2 averaged −19.3±0.8 (n=11), −18.8±1.4 (n=12), and −17.4±0.5 (n=12) mV in P0, P1, and P3 cells, respectively (P=NS). Biexponential ICaL inactivation time constants were unchanged by pacing (Figure 2D). Time-dependent recovery was well fitted by monoexponential functions, with time constants averaging 46.9±2.4 ms (n=12 cells), 41.9±2.2 ms (n=13), and 42.7±4.1 ms (n=10) in P0, P1, P3 cells (P=NS; Figure 2E).
Recordings used to obtain low-voltage-activated T-type calcium current (ICaT) are shown in supplemental Figure IIA. As previously described,4 ICaT was analyzed by subtracting currents recorded with a holding potential of −90 mV from currents in the same cell at a holding potential of −50 mV. Mean ICaT density–voltage relations are shown in supplemental Figure IIB. ICaT density was unaffected by changing pacing frequency during the 24-hour culture period. For example ICaT at −20 mV averaged −1.30±0.1 pA/pF (n=15), −1.25±0.1 pA/pF (n=20), and −1.28±0.1 pA/pF (n=19) in P0, P1, and P3 cells, respectively (P=NS). Current–voltage relations showed no differences among groups, suggesting no differences in voltage dependence.
Transient Outward Current
Ito recordings from P0, P1, and P3 cells are shown in supplemental Figure IIIA. Ito was significantly reduced by tachypacing (supplemental Figure IIIB). For example, Ito at +40 mV averaged 7.6±0.5 (n=20) pA/pF, 7.6±0.6 pA/pF (n=18), and 3.5±0.4 pA/pF (n=19) in P0, P1, and P3 cells, respectively. The overall form of the current–voltage relation did not change. The approach used in Figure 2C was applied to determine Ito activation voltage dependence, with the Ito reversal potential based on tail currents following 2.2-ms activating pulses to +50 mV averaging −71±2 mV, −73±4 mV, −70±2 mV (n=5 cells per group) in P0, P1, P3 cells (P=NS). Activation V1/2 based on data fits in each experiment (supplemental Figure IIIC) averaged 9.5±1.6 (n=17), 9.4±1.1 (n=15) and 10.7±1.9 mV (n=13) in P0, P1, and P3 cells (P=NS). Inactivation voltage dependence was studied with 1000-ms prepulses from −70 mV, followed by 200-ms test pulses to +60 mV. Boltzmann fit inactivation V1/2 averaged −24.9±1.2 (n=22 cells), −26.7±0.9 (n=21), and −26.7±1.4 mV (n=19) in P0, P1, and P3 (P=NS) cells (supplemental Figure IIIC). The time constants of Ito inactivation curves, which were well fitted by monoexponential relations, showed no differences (supplemental Figure IIID).
Inward Rectifier and Delayed Rectifier K+ Currents
Supplemental Figure IIC shows 1 mmol/L Ba2+-sensitive IK1 density–voltage relations, which were comparable among P0, P1, and P3 cells. Supplemental Figure IID and IIE shows current–voltage relations for E-4031–resistant IKs and HMR1566-resistant IKr tail currents in P0, P1, and P3 cells. There were no significant differences in IKr or IKs of P0, P1, and P3 cells.
Intracellular Calcium Responses
The results presented in Figures 1 and 2⇑ and supplemental Figures II and III suggest that the isolated cell model recapitulates the principal cellular electrophysiological features of atrial tachycardia remodeling: reduced APD and APD rate adaptation, decreased ICaL and Ito, and unchanged ICaT, IKr, and IKs.4 A commonly accepted explanation of the mechanism of atrial tachycardia remodeling invokes a negative-feedback response to protect against Ca2+ overload via the downregulation of ICaL.2 However, no direct evidence to support this notion has been presented. Our system allows for the measurement of [Ca2+]i and ICaL as a function of in vitro pacing duration, permitting an evaluation of the relative changes in each component. We therefore recorded [Ca2+]i and ICaL in parallel sets of cells from the same batches that were either not tachypaced (P1) or were paced at 3 Hz (P3) for 2, 8, or 24 hours. Figure 3A shows examples of [Ca2+]i transients measured at 1 Hz from cells that had been subjected to 3-Hz pacing for 8 and 24 hours. Both diastolic and systolic [Ca2+]i were greatly increased at 8 hours and returned toward baseline values after 24 hours. Mean peak systolic and diastolic [Ca2+]i are shown in Figure 3B and 3C and indicate statistically significant increases with 3-Hz pacing versus prepacing baseline and P1 cells at 2 and 8 hours, with a return to values not significantly different from baseline at 24 hours. The [Ca2+]i transient decay time constant decreased at 2 and 8 hours, consistent with accelerated removal of Ca2+ as an adaptation to Ca2+ loading. The lack of change over time in P1 cells indicates that the [Ca2+]i transient responses in P3 cells are a function of tachypacing and not simply the result of pacing in culture. Associated changes in ICaL are illustrated in Figure 3E through 3G. ICaL recordings after various tachypacing durations are shown in Figure 3E. Mean ICaL remained unchanged during up to 8 hours of tachypacing but decreased significantly at 24 hours (Figure 3F). ICaL inactivation kinetics (supplemental Figure IVA) and voltage dependence (supplemental Table II) were not affected by tachypacing. The time course of peak ICaL changes is illustrated in Figure 3G, and indicates that reductions in ICaL correspond temporally to the restoration of baseline [Ca2+]i values. No changes in ICaL were observed over time in P1 cells (supplemental Figure V). These results provide evidence for a molecular feedback loop that involves cellular sensing of [Ca2+]i loading, coupling to regulatory mechanisms that reduce ICaL and Ca2+ loading, resulting in a restoration of cellular Ca2+ balance. We therefore turned our attention to the intracellular signaling mechanisms that sense Ca2+ loading and cause ICaL downregulation.
Mechanisms of ICaL Downregulation
We first addressed the mechanism for ICaL downregulation. There is evidence that in vivo atrial tachycardia downregulates ICaL via decreases in α1c subunit mRNA expression.4,15 We therefore applied QPCR to measure Cav1.2 mRNA expression in P1 and P3 cells subjected to various pacing durations. Cav1.2 mRNA expression was significantly decreased, by 32±4% and 48±4%, respectively, after 8 and 24 hours of 3-Hz pacing (Figure 4A, right). Cav1.2 protein expression was not significantly altered at 2 or 8 hours but was reduced significantly at 24 hours (Figure 4B; examples of Western blots in supplemental Figure VI), by 47±8%, in P3 cells. The changes in Cav1.2 mRNA and protein expression were specifically related to tachypacing because neither was altered significantly in P1 cells (Figure 4, left).
Analysis of Ca2+-Dependent Pathways
The data described above point to a Ca2+-dependent pathway that is triggered by increased [Ca2+]i, leading to reductions in Cav1.2 mRNA within 8 hours, which results in decreased Cav1.2 protein expression and diminished ICaL within 24 hours. If this presumed system is correct, prevention of intracellular Ca2+ loading in response to 3-Hz pacing should prevent ICaL downregulation. We tested this idea in 2 ways. First, we exposed P0, P1, and P3 cells throughout 24-hour pacing to 1 μmol/L nimodipine to suppress Ca2+ entry via ICaL. Consistent with the hypothesized role of Ca2+ entry in ICaL downregulation, nimodipine prevented ICaL decreases in P3 cells (Figure 5A). In additional studies, we added 10 μmol/L BAPTA-AM to the culture medium for the 24-hour-pacing period to buffer [Ca2+]i. Once again, ICaL downregulation was prevented in P3 cells (Figure 5B). Nimodipine and BAPTA-AM also controlled intracellular Ca2+ loading, as indicated by Ca2+ imaging (supplemental Figure VII).
We then turned our attention to potential Ca2+-dependent signaling pathways. Calmodulin is a key Ca2+-binding protein that senses intracellular Ca2+ concentration and modulates a wide range of Ca2+-dependent enzyme systems.16 We used the calmodulin inhibitor W-7 to address the potential importance of calmodulin in coupling [Ca2+]i increases to downstream effectors. Atrial cardiomyocytes were subjected to P0, P1, or P3 conditions for 24 hours in the presence of 1 μmol/L W-7. Calmodulin inhibition completely suppressed ICaL downregulation in P3 cells (Figure 5C). Furthermore, calmodulin inhibition prevented downregulation of Cav1.2 transcript expression by tachypacing (supplemental Figure VIIIA).
Calcineurin is a Ca2+-dependent protein phosphatase that plays a key role in a variety of cardiac-remodeling processes,17 including the regulation of ion channel expression.18 We studied potential calcineurin involvement by adding the calcineurin blocker FK-506 (5 μmol/L) to the medium of P0, P1, and P3 cells during 24-hour pacing. FK-506 prevented ICaL downregulation (Figure 5D), as well as tachypacing-induced Cav1.2 transcript downregulation (supplemental Figure VIIIB). We then measured calcineurin-related phosphatase activity over time in cells exposed to 1- or 3-Hz pacing. No significant changes in calcineurin activity were observed in P1 cells (Figure 4C). In contrast, calcineurin activity was increased significantly at 2 and 8 hours of tachypacing in P3 cells. This time course indicates that calcineurin activity is rapidly increased by intracellular Ca2+ loading and that calcineurin activity increases precede decreases in Cav1.2 mRNA, consistent with a calcineurin-dependent signal that causes downregulation in mRNA expression.
Calcineurin dephosphorylates cytoplasmic NFAT transcription factors, promoting their translocation into the nucleus, where they participate in transcriptional regulation.19 We used deconvolved confocal microscopic immunofluorescent images to quantify cellular localization of the NFAT isoforms NFATc3 and NFATc4. Figure 6A shows examples of NFATc3, NFATc4, and merged images (with ToPro3 to define the nucleus) in P1 and P3 cells. Figure 6B shows quantification of the nuclear/cytoplasmic staining ratios. Whereas P1 and P0 cells showed similar nuclear/cytoplasmic staining, P3 cells showed stronger nuclear staining localization, consistent with nuclear translocation. If NFAT-dependent transcription changes are important in ICaL downregulation, NFAT inhibition should suppress tachypacing-induced ICaL downregulation. We, therefore, added the cell-permeable NFAT inhibitor INCA-6 (10 μmol/L) to the culture medium during 24-hour pacing. Consistent with an important role for calcineurin-dependent NFAT changes, INCA-6 completely prevented ICaL downregulation (Figure 5E). For INCA-6, as for all of the interventions we studied, neither ICaL inactivation kinetics (supplemental Figure IV) nor voltage dependence (supplemental Table II) was altered with cell pacing. Consistent with nuclear NFAT translocation playing a key downstream role in Cav1.2/ICaL downregulation, interventions acting higher up in the signaling cascade (BAPTA-AM, W-7, and FK-506) all suppressed NFAT translocation (Figure 6C through 6E). Interestingly, despite the fact that W-7 and FK-506 prevented transcriptional downregulation of ICaL, they did not adversely affect cell Ca2+ indices; on the contrary, they prevented Ca2+ transient increases with 8-hour P3 pacing (supplemental Figure IX).
Prevention of APD Changes
One potential interest of delineating the signal transduction processes involved in rate-dependent remodeling would be to design new strategies to prevent arrhythmogenic APD shortening. As a proof of principle, we studied the effects on cultured-myocyte APD of inhibiting various components of the signaling system. Figure 7A shows APs recorded at 1 Hz from P1 and P3 cells exposed to nimodipine, BAPTA-AM, W-7, or FK-506 throughout the 24-hour pacing period. Figure 7B shows corresponding mean±SEM data for P0, P1, and P3 cells. The results indicate that prevention of Ca2+ loading, suppression of calmodulin signaling, or inhibition of calcineurin prevents tachycardia-induced APD abbreviation. There might be concern that prevention of ICaL downregulation could have negative effects on cell viability by enhancing Ca2+ overload. On the contrary, interventions that prevented ICaL downregulation were found also to prevent the deterioration in cell viability (supplemental Table I) and Ca2+ overload (supplemental Figure VII and IX) caused by tachypacing.
In the present study, we explored the cell-signaling that causes APD abbreviation and ICaL downregulation in an in vitro model of atrial tachycardia remodeling. The results suggest that ICaL downregulation results from a molecular feedback system that detects intracellular Ca2+ loading via the Ca2+/calmodulin system, which activates calcineurin, which, in turn, stimulates nuclear translocation of NFAT, resulting in transcriptional downregulation of Cav1.2 that decreases Ca2+ load by reducing ICaL, with APD reduction as a consequence.
Comparison With Previous Studies of Atrial Tachycardia Remodeling
Decreased ICaL is a consistent feature of atrial tachycardia remodeling and is believed to contribute importantly to the APD abbreviation that shortens refractoriness and promotes AF.4,6,20 ICaL reduction is associated with decreased Cav1.2 mRNA expression,5,10,15,21 and many investigators have also noted decreases in Cav1.2 protein expression.10,21,22 A role for Ca2+ loading in atrial tachycardia remodeling was first suspected based on the protective effects of ICaL blockers against short-term remodeling.8,23 Subsequent work showed that atrial cardiomyocyte Ca2+ loading begins within several minutes of the onset of cellular tachypacing.24 Cell ultrastructure changes point to transient Ca2+ overload in atria of goats with AF.9 Short-term tachypacing of rat atria augments nuclear NFAT expression and causes changes in BNP and c-fos expression, effects that are prevented by cyclosporin-A.19 Bukowska et al showed calcineurin expression and activity upregulation in atrial tissues from AF patients, in association with nuclear translocation of NFAT and increased hypertrophic gene expression.25 Calcineurin-related changes were reproduced by ex vivo pacing (2 to 4 Hz) of human atrial tissue slices.25 Lin et al presented additional evidence for in vivo activation of the calcineurin and NFAT pathways in an atrial tachypacing porcine AF model.26
These previous studies suggest that atrial tachycardia causes Ca2+ loading, induces transcriptional downregulation of Cav1.2, and can activate calcineurin-dependent mechanisms. Our study is the first to demonstrate directly causative links between atrial cardiomyocyte tachycardia, cellular Ca2+ loading, ICaL downregulation, calcineurin activation, NFAT translocation, and associated APD alterations. A schematic representing our findings is presented in Figure 8. The solid vertical arrows indicate directly measured changes in [Ca2+]i, Cav1.2 mRNA and protein, ICaL, calcineurin activity, nuclear NFAT localization, and APD. Experiments with selective inhibitors (indicated in boxes) confirm important mediating roles for free [Ca2+]i increases, calmodulin, calcineurin, and NFAT. Our time course studies suggest the presence of an intracellular molecular feedback loop initiated by [Ca2+]i accumulation, which leads to downregulation of Cav1.2 mRNA, followed by decreased Cav1.2 protein expression, which downregulates ICaL and mitigates [Ca2+]i loading.
Ca2+/Calmodulin and Calcineurin-Mediated Ion Channel Regulation
The Ca2+/calmodulin system is an important cardiomyocyte Ca2+ sensor that responds to intracellular Ca2+ changes on both a beat-to-beat and tonic basis.27 Ca2+/calmodulin binding activates a variety of downstream mediators, including calcineurin and Ca2+/calmodulin kinase II. Neonatal rat cardiomyocyte tachystimulation activates calcineurin/NFAT signaling.28 Calcineurin is a Ca2+-activated serine/threonine phosphatase composed of a catalytic A subunit (56 to 63 kDa) and a regulatory B subunit (19 kDa)17 that dephosphorylates a number of cytoplasmic proteins, including the regulatory domains of NFATc3 and NFATc4, which translocate into the nucleus and regulate gene transcription.29,30 NFAT is a particularly important mediator of responses to changes in intracellular Ca2+ patterns.31 Calcineurin/NFAT signaling alters transcriptional regulation of Ito, with most studies showing Ito downregulation,18,32–34 although upregulation has also been observed in neonatal rat cardiomyocytes.35 ICaL is increased in calcineurin-overexpressing mice, but this effect appears to be related to hypertrophy rather than to direct effects on the ICaL system.36 We did not observe hypertrophic responses in tachystimulated atrial cardiomyocytes.
Considerations of the Model
Our model reproduced many features of atrial tachycardia remodeling observed in atrial tissue samples obtained from AF patients or tachypaced animals such as dogs, pigs, and sheep, including reduced APD and APD rate dependence; reduced ICaL and Ito; and unchanged ICaT, IKr, and IKs.2,37 Previous in vitro models of atrial tachycardia remodeling have included rapidly paced HL-1 cells derived from a mouse atrial tumor cell line,38 and paced human atrial tissue slices.25 The HL-1 cell model displays important features of atrial remodeling but is limited by a low expression rate of ICaL (<20% of cells display prominent ICaL) and K+ channel remodeling (unchanged Ito, increased IKr)38 that differs from the established pattern of in vivo atrial tachycardia remodeling (reduced Ito, unchanged IKr).4,20,37 The human tachypaced atrial slice model reproduces some biochemical features seen in AF patients, but its electrophysiological properties have not been established, and atrial tissue superfusion is known to cause important functional alterations and viability problems.39 One potentially important change that we did not observe in tachypaced cardiomyocytes was inward rectifier K+ current enhancement.37 A possible explanation for this is that inward rectifier K+ current enhancement develops more slowly than the other changes, not having been noted with <7-day atrial tachyarrhythmia.40
Tachypaced cardiomyocytes in our model did not show cellular hypertrophy, consistent with previous studies of atrial cells from dogs subjected to up to 42-day tachypacing.4 Atrial cells from patients with persistent AF may be hypertrophied,6 possibly because of underlying cardiac disease or the effects of greater AF chronicity, a difference that must be considered in interpreting our findings.
Potential Significance of Our Observations
To our knowledge, our study is the first to explore in detail the molecular signaling underlying atrial tachycardia cellular electrophysiological remodeling. Atrial tachycardia remodeling has significant clinical consequences.2,3 Suppression of atrial remodeling is a promising target for AF therapy development,7,41 but a major limitation to therapeutic innovation in this area is a lack of clear information about underlying molecular mechanisms. The cellular signaling underlying atrial tachycardia remodeling defined in the present study is, therefore, of considerable potential significance. Furthermore, we have shown that inhibition of key steps in the signaling chain can prevent remodeling-induced AP changes that are important in AF promotion. Interestingly, interventions that suppressed ICaL downregulation improved Ca2+i loading and cell viability, suggesting that the signaling cascade producing ICaL downregulation may not effectively prevent Ca2+ overload and cell damage and may actually produce counterproductive consequences. This is a novel observation with potentially important implications for the management of AF and the development of antiremodeling therapies. In addition to identifying potential targets for antiremodeling therapy, we have established a cellular model of atrial tachycardia remodeling that may be useful for further investigations of the underlying cellular and molecular biology.
We thank Chantal St-Cyr and Nathalie L’Heureux for technical assistance, France Thériault for secretarial support, and Balazs Ordog for help with QPCR.
Sources of Funding
Supported by Canadian Institutes of Health Research (MOP 44365); Quebec Heart Foundation; Dutch Heart Foundation (2007B217); Dutch Organization for Scientific Research (NWO program grant 916.46.043); German Federal Ministry of Education and Research (Atrial Fibrillation Competence Network grant 01Gi0204); and a network grant (European–North American Atrial Fibrillation Research Alliance) from Fondation Leducq.
Original received March 11, 2008; revision received July 10, 2008; accepted August 13, 2008.
Benjamin EJ, Wolf PA, D'Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation. 1998; 98: 946–952.
Wijffels MC, Kirchhof CJ, Dorland R, Power J, Allessie MA. Electrical remodeling due to atrial fibrillation in chronically instrumented conscious goats: roles of neurohumoral changes, ischemia, atrial stretch, and high rate of electrical activation. Circulation. 1997; 96: 3710–3720.
Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes in a canine model of atrial fibrillation. Circ Res. 1997; 81: 512–525.
van der Velden HMW, van der Zee L, Wijffels MC, van Leuven C, Dorland R, Vos MA, Jongsma HJ, Allessie MA. Atrial fibrillation in the goat induces changes in monophasic action potential and mRNA expression of ion channels involved in repolarization. J Cardiovasc Electrophysiol. 2000; 11: 1262–1269.
Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial L-type Ca2+ currents and human atrial fibrillation. Circ Res. 1999; 85: 428–436.
Goette A, Honeycutt C, Langberg JJ. Electrical remodeling in atrial fibrillation. Time course and mechanisms. Circulation. 1996; 94: 2968–2974.
Yue L, Melnyk P, Gaspo R, Wang Z, Nattel S. Molecular mechanisms underlying ionic remodeling in a dog model of atrial fibrillation. Circ Res. 1999; 84: 776–784.
Zhang LM, Wang Z, Nattel S. Effects of sustained beta-adrenergic stimulation on ionic currents of cultured adult guinea pig cardiomyocytes. Am J Physiol. 2002; 282: H880–H889.
Coutu P, Chartier D, Nattel S. Comparison of Ca2+-handling properties of canine pulmonary vein and left atrial cardiomyocytes. Am J Physiol. 2006; 291: H2290–H2300.
Burstein B, Qi XY, Yeh YH, Calderone A, Nattel S. Atrial cardiomyocyte tachycardia alters cardiac fibroblast function: a novel consideration in atrial remodeling. Cardiovasc Res. 2007; 76: 442–452.
Brundel BJ, van Gelder IC, Henning RH, Tuinenburg AE, Deelman LE, Tieleman RG, Grandjean JG, van Gilst WH, Crijns HJ. Gene expression of proteins influencing the calcium homeostasis in patients with persistent and paroxysmal atrial fibrillation. Cardiovasc Res. 1999; 42: 443–454.
Molkentin JD. Calcineurin and beyond: cardiac hypertrophic signaling. Circ Res. 2000; 87: 731–738.
Rossow CF, Minami E, Chase EG, Murry CE, Santana LF. NFATc3-induced reductions in voltage-gated K+ currents after myocardial infarction. Circ Res. 2004; 94: 1340–1350.
Bosch RF, Zeng X, Grammer JB, Popovic K, Mewis C, Kuhlkamp V. Ionic mechanisms of electrical remodeling in human atrial fibrillation. Cardiovasc Res. 1999; 44: 121–131.
Brundel BJ, Van Gelder IC, Henning RH, Tieleman RG, Tuinenburg AE, Wietses M, Grandjean JG, Van Gilst WH, Crijns HJ. Ion channel remodeling is related to intraoperative atrial effective refractory periods in patients with paroxysmal and persistent atrial fibrillation. Circulation. 2001; 103: 684–690.
Klein G, Schroder F, Vogler D, Schaefer A, Haverich A, Schieffer B, Korte T, Drexler H. Increased open probability of single cardiac L-type calcium channels in patients with chronic atrial fibrillation. Role of phosphatase 2A. Cardiovasc Res. 2003; 59: 37–45.
Leistad E, Aksnes G, Verburg E, Christensen G. Atrial contractile dysfunction after short-term atrial fibrillation is reduced by verapamil but increased by BAY K8644. Circulation. 1996; 93: 1747–1754.
Sun H, Chartier D, Leblanc N, Nattel S. Intracellular calcium changes and tachycardia-induced contractile dysfunction in canine atrial myocytes. Cardiovasc Res. 2001; 49: 751–761.
Xia Y, McMillin JB, Lewis A, Moore M, Zhu WG, Williams RS, Kellems RE. Electrical stimulation of neonatal cardiac myocytes activates the NFAT3 and GATA4 pathways and up-regulates the adenylosuccinate synthetase 1 gene. J Biol Chem. 2000; 275: 1855–1863.
Hogan PG, Chen L, Nardone J, Rao A. Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev. 2003; 17: 2205–2232.
Dong D, Duan Y, Guo J, Roach DE, Swirp SL, Wang L, Lees-Miller JP, Sheldon RS, Molkentin JD, Duff HJ. Overexpression of calcineurin in mouse causes sudden cardiac death associated with decreased density of K+ channels. Cardiovasc Res. 2003; 57: 320–332.
Perrier E, Perrier R, Richard S, Benitah JP. Ca2+ controls functional expression of the cardiac K+ transient outward current via the calcineurin pathway. J Biol Chem. 2004; 279: 40634–40639.
Rossow CF, Dilly KW, Santana LF. Differential calcineurin/NFATc3 activity contributes to the Ito transmural gradient in the mouse heart. Circ Res. 2006; 98: 1306–1313.
Gong N, Bodi I, Zobel C, Schwartz A, Molkentin JD, Backx PH. Calcineurin increases cardiac transient outward K+ currents via transcriptional up-regulation of Kv4.2 channel subunits. J Biol Chem. 2006; 281: 38498–38506.
Nattel S, Maguy A, Le Bouter S, Yeh Y-H. Arrhythmogenic ion-channel remodeling in the heart: heart failure, myocardial infarction, and atrial fibrillation. Physiol Rev. 2007; 87: 425–456.
Burashnikov A, Mannava S, Antzelevitch C. Transmembrane action potential heterogeneity in the canine isolated arterially perfused right atrium: effect of IKr and IKur/Ito block. Am J Physiol Heart Circ Physiol. 2004; 286: H2393–H2400.
Voigt N, Friedrich A, Bock M, Wettwer E, Christ T, Knaut M, Strasser RH, Ravens U, Dobrev D. Differential phosphorylation-dependent regulation of constitutively active and muscarinic receptor-activated IK,ACh channels in patients with chronic atrial fibrillation. Cardiovasc Res. 2007; 74: 426–437.