Cellular Mechanisms of Atrial Contractile Dysfunction Caused by Sustained Atrial Tachycardia
Background—Transient atrial contractile dysfunction (“atrial stunning”) follows conversion of atrial fibrillation (AF) to sinus rhythm and has significant clinical implications; however, the underlying mechanisms are poorly understood. We investigated the hypothesis that rapid atrial activation (as during AF) impairs cellular contractility and affects cellular Ca2+ handling.
Methods and Results—Edge detection and indo 1 fluorescence techniques were used to measure unloaded cell shortening and intracellular Ca2+ transients in atrial myocytes from control (Ctl) dogs and dogs subjected to atrial pacing at 400 bpm for 7 (P7) or 42 (P42) days. Atrial tachycardia reduced fractional cell shortening (0.1 Hz) from 7.3±0.4% (Ctl) to 4.3±0.3% and 2.0±0.3% in P7 and P42 dogs, respectively (P<0.01 for each). Resting [Ca2+]i was not altered in paced dogs, but the systolic Ca2+ transient was significantly reduced. Furthermore, cells from paced dogs showed slowed relaxation and use-dependent decreases of Ca2+ transients and cell shortening compared with cells from Ctl dogs. To determine whether changes in Ca2+ transients account fully for alterations in contractility, we varied [Ca2+]o to evaluate the relation between Ca2+ transients and cell shortening. Reductions in Ca2+ transients in Ctl cells reduced shortening to the level of paced cells; however, when Ca2+ transients in P42 cells were elevated to the range of Ctl cells, a significant reduction in cell shortening remained. Similar results were obtained in dogs that maintained 1:1 capture throughout the monitoring period and dogs that developed sustained AF over the course of the study.
Conclusions—Sustained atrial tachycardia causes important reductions in cellular contractility, in part by impairing cellular Ca2+ handling and decreasing systolic Ca2+ transients. These results provide direct evidence for the concept that AF induces atrial contractile dysfunction by causing a tachycardia-induced atrial cardiomyopathy.
A delay in the return of atrial mechanical function after cardioversion of AF to sinus rhythm has been well recognized since the 1960s1 2 3 4 5 6 and most likely accounts for a delayed improvement of functional capacity.7 Recently developed pulsed Doppler techniques have shown that several weeks may be required for full atrial functional recovery after cardioversion8 9 10 11 and that the degree of contractile dysfunction and the time required for recovery are related to the duration of AF.9 10 11 Contractile dysfunction is mild after short-duration AF and recovery of atrial function occurs rapidly after cardioversion, whereas chronic AF produces severe dysfunction that may require up to 1 month to achieve maximal recovery.9 10 11 Prospective studies have confirmed a delayed restoration (days to weeks) of atrial contractile function after the return of normal electrical activity in almost all patients with chronic AF.7 8 9 10 11 Transient atrial dysfunction may be important in the process of atrial stunning, which is believed to play a role in thromboembolic complications among AF patients after rhythm reversion.12 13 14
The mechanisms of atrial contractile dysfunction after conversion of AF are poorly understood. Leistad et al15 showed that even short-term AF (5 minutes) was followed by impaired atrial contractility in pigs. Atrial dysfunction was reduced by exposure to the Ca2+ antagonist verapamil and increased by the Ca2+ agonist Bay K8644, suggesting that Ca2+ overload may play a role. Atrial ischemia did not appear to be involved. Morillo et al16 showed that rapid atrial pacing (400 bpm) for 42 days causes atrial dilation and permits the induction of sustained AF. These findings raise the possibility that the chronic rapid atrial activation of AF induces an atrial cardiomyopathy much like the well-recognized ventricular cardiomyopathy caused by chronically elevated ventricular rates.17 We have shown that atrial pacing at 400 bpm causes time-dependent reductions in the transient outward K+ current (Ito) and L-type Ca2+ current (ICa), with the latter appearing to be particularly important in causing action potential changes associated with the ability to maintain sustained AF.18 Because ICa is important in providing the trigger for SR Ca2+ release and in maintaining the loading state of SR Ca2+ stores,19 abnormalities in ICa and/or other elements of Ca2+ handling that result from atrial tachycardia may cause contractile dysfunction. The present study was designed to use video edge-detection and microfluorimetric cellular Ca2+ measurements to determine (1) whether sustained atrial tachycardia causes impaired cellular contractility, (2) whether the systolic Ca2+ transient is altered in atrial myocytes from dogs subjected to chronic rapid atrial pacing, and (3) whether abnormal Ca2+ handling in itself can explain the cellular contractile dysfunction in rapidly paced dogs.
Preparation of the Animal Model
Adult mongrel dogs of either sex (25 to 30 kg) were anesthetized with sodium pentobarbital (30 mg/kg IV, followed by boluses of 4 mg/kg IV as needed). Artificial respiration was maintained via an endotracheal tube connected to a mechanical ventilator. Under sterile conditions, a unipolar screw-in Medtronic J pacing lead (Medtronic Inc) was inserted through the right jugular vein, and the distal end was fixed in the right atrial appendage under fluoroscopic guidance. The proximal end of the lead was connected to an implantable Medtronic pacemaker unit (model 8084) inserted into a subcutaneous pocket in the neck. After 24 hours for lead stabilization, the pacemaker was programmed to capture the atria at 400 bpm (150-ms cycle length) with 4-ms square-wave pulses at twice threshold current. The right atrium was stimulated at this rate for 7 (P7 dogs, n=6) or 42 (P42 dogs, n=6) days. The surface ECG was verified after 24 hours and then weekly to ensure atrial capture. Dogs who developed AF were not cardioverted; they were left in AF until the end of the study. Rapidly paced dogs were compared with a Ctl group (n=9), including 3 sham-operated dogs (pacemaker inserted but not activated) monitored for 7 days.
Cell Isolation and Solutions
On study days, dogs were anesthetized with morphine (2 mg/kg SC) and α-chloralose (120 mg/kg IV) and ventilated with room air supplemented with oxygen. A median sternotomy was performed, and a surface ECG and (in paced dogs) an atrial electrogram were recorded. The implanted pacemaker was then deactivated and the heart rapidly removed after an intra-atrial injection of 10 000 U heparin. The heart was immersed in Ca2+-containing Tyrode’s solution aerated with 100% O2, the left circumflex coronary artery was cannulated, and the left atrium was perfused with Ca2+-containing Tyrode’s solution until clear of blood. The perfusate was then changed to nominally Ca2+-free Tyrode’s solution for 20 minutes, followed by nominally Ca2+-free Tyrode’s solution containing 100 U/mL collagenase (type 2, Worthington) and 0.1% BSA for 60 to 80 minutes (actual perfusion time determined by time for maximum cell yield). All perfusion solutions were aerated with 100% O2 and warmed to 37°C. The tissue was then placed in a Tyrode’s storage solution containing 10 μmol/L Ca2+, and the cells were dispersed by gentle trituration. The concentration of Ca2+ in the cellular supernatant was then increased to 100 μmol/L by addition of a small volume of storage solution containing 1 mmol/L Ca2+. The cells were kept in this solution at room temperature for use within 12 hours after isolation.
The Tyrode’s solution contained (in mmol/L) NaCl 136, KCl 5.4, MgCl2 1.0, CaCl2 2.0, NaH2PO4 0.33, glucose 10, and HEPES 10, pH adjusted to 7.4 with NaOH. The nominally Ca2+-free Tyrode’s solution had the same composition except that CaCl2 was omitted. The storage solution contained (in mmol/L) NaCl 136, KCl 5.0, MgCl2 1.0, NaH2PO4 0.33, glucose 10, HEPES 10, and CaCl2 0.01 or 1.0, plus BSA 0.2%, pH adjusted to 7.35 with NaOH. The normal bath solution for experiments was the same as Tyrode’s solution except for the CaCl2 concentration (1.8 mmol/L). In some experiments, Tyrode’s solution modified to contain low (0.3 or 0.9 mmol/L) or high (2.7 or 5.4 mmol/L) Ca2+ concentrations were used as specified below.
Field Stimulation and Measurement of Cell Contraction
Isolated myocytes were field-stimulated via 10-ms square-wave pulses with 1.5 times threshold amplitude. Stimuli were delivered by a stimulator (S48, Grass Instruments) via a platinum bipolar electrode mounted on a micromanipulator. CS was measured with a video edge detector (Crescent Electronics) coupled to a charge-coupled camera mounted on the side port of the microscope. The contraction signal was sampled at 200 Hz (TL-1 A/D Converter, Axon) and delivered to an IBM-compatible computer via Axotape software (Axon). The cursors for edge detection were generally positioned at both ends of the cell to measure whole-cell shortening. Examples for 1 cell of microscopic images, along with cursor placement and the corresponding length recording, are shown in Figure 1⇓.
Intracellular CaTs were recorded by a fluorimetric ratio technique (indo 1 fluorescence) similar to that previously described.20 The fluorescent indicator indo 1 was loaded by incubating the myocytes at room temperature for 10 to 12 minutes with 5 μmol/L of the acetoxymethyl ester form (indo 1-AM, Molecular Probes) in 100 μmol/L Ca2+ storage Tyrode’s solution. Myocytes were then perfused with normal bath solution at room temperature for at least 40 minutes to wash out extracellular indicator and to allow for intracellular deesterification of indo 1. The loading procedure did not significantly alter CS (Table⇓) but reduced maximal shortening and relaxation rates (+dL/dt and −dL/dt) in Ctl cells. Consequently, contractility was analyzed primarily on the basis of studies without indo 1 loading, with CS monitored during indo 1 studies only to verify qualitatively comparable contractile behavior of the cells used to study CaTs. Background and cell autofluorescence were cancelled out by zeroing the output of the photomultiplier tubes using cells without indo 1 loading.
In this study, R400/500 was used as the index of [Ca2+]i as previously described.20 This approach avoids uncertainties related to in vivo calibration of fluorescent Ca2+ indicators.21 Ultraviolet light emanating from a mercury arc lamp was used to excite indo 1. Exposure of the cell to UV light was controlled by an electronic shutter (Optikon, model T132, Vincent Associates) anchored between the arc lamp and epifluorescence attachment of an inverted Nikon Diaphot epifluorescence microscope (×40 Nikon oil-immersion fluor objective; numeric aperture, 1.3). The shutter could be manually triggered in the open position to verify the level of loading of indo 1 or to perform actual [Ca2+]i measurements. Only a portion of the cell was exposed to UV light (a circular beam ≈15 μm in diameter). The dye was excited at 340 nm by means of a narrow-bandwidth filter (±10 nm) and a dichroic mirror (>380 nm). The emitted fluorescent light (>380 nm) was then relayed to the lateral port of the microscope and processed by a spectral microfluorometer (Sycamore Scientific) equipped with a charge-coupled camera (Pulnix America Inc, model TM-440) and a TV monitor (JVC, model TM-122U) to view the image of the cell throughout the experiment and ensure its proper alignment within the detection window. The emitted light was first split by a series of dichroic mirrors and passed through narrow–band-pass (±10 nm) filters centered at 400 and 500 nm. Light intensity was monitored by means of 2 matched photomultiplier tubes (Hamamatsu type R2560HA). The analog ratio (Burr Brown Corp, model M/N DIV100HP) of the 2 fluorescent signals (400/500 nm) was electronically filtered with a low-pass Butterworth filter set at 60 Hz (Frequency Devices model 901) and digitized on line at 1 kHz by use of a computer-based acquisition system (TL-1–125 LabMaster Board, Axon Instruments Inc) and software (Axotape version 2.0, Axon Instruments Inc) to be stored on the computer’s hard disk (PC-486/33 MHz) for later analysis and display.
All experiments were performed at 35.5°C (TC-202 Temperature Controller, Medical Systems Corp). After cell perfusion with normal bath solution at room temperature for 30 (for experiments without indo 1 loading) or 40 (for cells loaded with the indicator) minutes, the bath temperature was gradually raised to 35.5°C, and cells were stimulated at 0.2 Hz. When CS reached a steady state, the stimulation was interrupted for 1 minute and then resumed at a test frequency until steady state was attained. To study frequency dependence, 5 stimulation frequencies (0.1, 0.2, 0.5, 1, and 2 Hz) were applied to each cell in a random fashion, continuously or with a 1-minute rest period between frequencies. Only 1 cell was studied for each cell aliquot added to the bath.
Data Analysis and Statistics
CS and CaT recordings were low-pass filtered by averaging 3 to 5 adjacent points before any analysis. The amplitude of CS (ΔL) was measured relative to diastolic cell length and then normalized to the resting cell length (Lmax) and expressed as percentage shortening (ΔL/Lmax). TTP CS was measured between the onset and the peak of contraction. Time for 50% relaxation (TR1/2) was measured from the peak of contraction to the time at which cell length returned by 50% to diastolic values. The maximal velocity of shortening (+dL/dt) and the maximal rate of relaxation (−dL/dt) were measured from time derivatives of the contraction signal. The CaT amplitude (ΔR400/500) was measured as the difference between diastolic and maximal systolic R400/500 ratio. TTP of CaT was measured as the period between the onset and the peak of the fluorescence signal. In cells exhibiting 2 rising phases of CaT, the time to the first peak was measured. The time constant of relaxation of the CaT (τR) was determined by nonlinear least-squares curve fitting (Chebyshev method in pClamp 6) over a 1-second period beginning from the peak of the CaT. Each variable was quantified on the basis of the average measurement for 5 consecutive beats at steady state.
Because the results of the 3 sham-operated dogs were not different from those of the 6 dogs without pacemaker insertion, their results were grouped together as a single control group. Among the 6 P42 dogs studied, 3 developed sustained AF during the course of the study, and 3 maintained 1:1 atrial capture at 400 bpm. There were no appreciable differences in the properties of CS (eg, at 0.1 Hz, ΔL/Lmax averaged 1.98±0.32% in 13 cells from dogs with AF versus 2.36±0.29% in 20 cells from dogs with 1:1 capture, P=NS) and CaTs (eg, ΔR400/500 averaged 0.43±0.04 in 11 cells from AF dogs and 0.46±0.03 in 10 cells from dogs with 1:1 capture, P=NS) between P42 dogs who developed AF and those in paced atrial tachycardia; therefore, their results were grouped together for some analyses. For the primary analyses of CS and CaT properties (Figures 2⇓ and 3⇓), results for ST dogs and AF dogs are presented separately.
The statistical significance of group differences was evaluated by ANOVA. The significance of differences between individual means was then examined with paired or unpaired Student’s t tests with Bonferroni’s correction. A 2-tailed P<0.05 was considered to indicate significance. Group data are expressed as the mean±SEM.
Effects of Sustained Atrial Tachycardia on Cellular Contractility
Figure 2A⇑ shows representative recordings of steady-state CS at 0.1 Hz from a Ctl, a P7, and a P42 cell. The amplitude of CS (ΔL/Lmax) is clearly decreased in the P7 cell and further reduced in the P42 cell. Overall, ΔL/Lmax was decreased by ≈40% in P7 cells (n=16) compared with Ctl (n=22) and was reduced by ≈70% in P42 cells, with similar results for cells from ST dogs (n=20) and from dogs that developed AF (n=13, Figure 1B⇑, P=NS for P42 ST versus AF cells). The maximum velocity of CS (Figure 2C⇑) and relaxation (Figure 2D⇑) decreased progressively in cells from paced dogs. No significant differences were noted in P42 dogs that developed chronic AF compared with those that maintained 1:1 atrial capture. These results indicate that chronic rapid activation reduced the extent of CS and decreased the rate of both contraction and relaxation.
Effects of Sustained Atrial Tachycardia on Intracellular Ca2+ Signals
To understand the mechanisms underlying the contractile dysfunction observed in myocytes isolated from P7 and P42 dogs, we examined the effects of rapid atrial pacing on the CaT, an important determinant of cardiac contractility. Figure 3A⇑ shows representative CaT recordings from Ctl, P7, and P42 cells loaded with indo 1 and stimulated at 0.1 Hz. Resting R400/500 levels (numbers to left of recordings) were similar in these cells; however, the amplitude of the CaT was slightly reduced in the P7 cell and substantially reduced in the P42 cell. Overall, the mean CaT (Figure 3C⇑) was reduced by 19% in 26 P7 cells compared with results in 27 Ctl cells (P<0.01) and further reduced in P42 cells (P<0.01 versus Ctl), with a similar reduction for ST dogs (n=10) and dogs that developed AF (n=11). Resting Ca2+ levels were not different among groups (Figure 3D⇑). The kinetics of the CaT were also altered by rapid pacing, as can be seen from the CaT recordings in Figure 3B⇑, in which 5 recordings from each of the cells shown in Figure 3A⇑ have been normalized to the maximum value of each recording to control for differences in amplitude and averaged. Both the rising and relaxation phases of the Ca2+ signal are slowed in paced cells, particularly from the P42 dog. As shown by mean data for the TTP of the CaT corrected for the CaT amplitude (TTP/ΔR400/500, Figure 3E⇑) and for the time constant of relaxation (τR, Figure 3F⇑), Ca2+ release and removal slowed progressively in cells from paced dogs. Thus, as for contractility, chronic tachycardia reduced the amplitude of Ca2+ release and slowed the kinetics of Ca2+ release and reuptake.
Effects of Sustained Atrial Tachycardia on Postrest Staircase Phenomena
Figure 4A⇓ shows representative recordings of CS (top) and CaT (bottom) obtained from a Ctl cell (left), a P7 cell (middle), and a P42 cell (right) stimulated at 1 Hz after a 1-minute rest period. Both the CS and CaT of the first postrest beat are larger than those at steady state, indicating postrest potentiation. All but 1 of 10 Ctl cells showed a progressive increase (positive staircase) in the amplitude of CS and the CaT after an initial decrease (negative staircase). In contrast, most P7 cells (10 of 12 cells) and all P42 cells (n=10) exhibited a monotonic and much more dramatic decrease in the amplitude of both CS and CaT during 1 minute of continued stimulation. Two P7 cells showed a small biphasic pattern, with a much lower amplitude of steady-state CS and CaT compared with Ctl cells. Mean data for CS (Figure 4B⇓) show a significant and similar postrest reduction in CS in P7 and P42 cells compared with Ctl and a much more marked decrease in steady-state CS. The results for CaTs (Figure 4B⇓, right) were qualitatively similar to those for CS, but the reductions in paced cells were not as striking. These observations suggest that rapid atrial activation causes abnormalities in time-dependent Ca2+ uptake, translocation, and/or release mechanisms, so that with repeated activation, Ca2+ cycling does not permit maintenance of the Ca2+ transient and the linked contraction process.
Effects of Sustained Atrial Tachycardia on the Frequency Dependence of Contraction and Ca2+Cycling
Figure 5⇓ shows representative recordings of steady-state CS (top) and CaTs (bottom) at 0.1, 0.5, and 1 Hz from a Ctl (top), a P7 (middle), and a P42 cell (bottom). At 0.1 Hz, both CS and the CaT were reduced in P7 and P42 cells compared with Ctl, as described above. When the stimulation frequency was increased, CS was slightly decreased in the Ctl cell (by ≈25%), and the CaT was unchanged. In P7 and P42 cells, however, increasing the stimulation frequency caused striking reductions in CS and the CaT. Figure 6⇓ shows mean values of CS (Figure 6A⇓) and the CaT (Figure 6B⇓) obtained at 5 stimulation frequencies for each cell for 9 Ctl, 14 P7, and 10 P42 cells. CS and CaTs were moderately reduced in paced cells at low frequencies, and reductions became much more dramatic at higher frequencies, with near–steady-state values reached at 0.5 Hz. Neither CS nor the CaT was significantly frequency dependent in Ctl cells; however, both decreased substantially in paced cells as frequency increased from 0.1 to 0.5 Hz, with an overall frequency dependence that was highly significant (P<0.001). Cells from dogs subjected to longer durations of pacing showed greater abnormalities at low frequencies, but as frequency increased, CS and the CaT were very strongly and similarly reduced relative to Ctl in both P7 and P42 cells.
Relationship Between Reductions in Contractility and Ca2+ Handling Caused by Sustained Atrial Tachycardia
The above results show that the alterations of CS and CaT caused by chronically rapid atrial activation are qualitatively similar in terms of kinetic changes during a beat, postrest behavior, and frequency dependence. We analyzed the relations between the CaT and CS to determine whether changes in contractility of paced cells can be explained fully by abnormalities in intracellular Ca2+ handling. We first determined the CS-CaT relation in a group of Ctl cells by changing the superfusate [Ca2+] from 0.3 to 2.7 mmol/L. This results in gradually increasing bath [Ca2+] over several minutes, thereby gradually increasing [Ca2+]o, until bath and superfusate [Ca2+] equilibrate. As bath [Ca2+] increases, numerous measurements of CS and CaT can be made simultaneously in each cell, defining the CS-CaT relation for that cell. Figure 7A⇓ shows graphs of CS versus the logarithm of the peak CaT measured at 0.1 Hz in each of 8 Ctl cells. CS increases with the CaT in each cell in a similar way. When all data from all 8 Ctl cells were analyzed together (Figure 7B⇓), the Ctl CS–versus–log CaT relation was well fit by a sigmoid relation, as classically described,22 providing the relation (solid curve) and 95% confidence limits (dashed curves) shown. The results obtained from P7 and P42 cells were then compared with the Ctl CS-CaT relationship, as shown in Figure 7C⇓ and 7D⇓, respectively. The solid symbols in these figures are results in Ctl cells (reproduced from Figure 7B⇓), and the open symbols are results for individual P7 and P42 cells at 1.8 mmol/L [Ca2+]o and 0.1 Hz. For P7 cells (Figure 7C⇓), the CaTs covered the entire range, and most values fell within the 95% confidence limits of Ctl cells. Although results from ≈75% of P42 cells (n=20) fell within the 95% confidence limits of the Ctl CS-CaT relation (Figure 7D⇓), the lack of cells with higher CaT values makes it difficult to know whether contractility would be normal in these cells if the CaT were normal, ie, whether all of their contractile abnormalities can be attributed to reduced activator Ca2+ release on stimulation.
To address this issue, we exposed a group of P42 cells (n=6) to gradually increasing bath Ca2+ concentrations from 1.0 to 5.4 mmol/L by increasing superfusate [Ca2+] from 1.0 to 5.4 mmol/L and making repeated measurements as bath [Ca2+] increased gradually over this range. The resulting CS-CaT relations are shown in Figure 8A⇓. The pattern of change in CS as a function of the CaT was quite constant among cells. Figure 8B⇓ shows the best-fit sigmoid curve to the CS-CaT relation of P42 cells (raw data from all cells shown by open symbols). For comparison, the results for Ctl cells (Figure 7B⇑) are reproduced by the solid symbols, along with the Ctl best-fit sigmoid relation. The results for P42 cells clearly fall below those of Ctl cells, indicating that even in the presence of the same CaT, P42 cells have an abnormal contractile response. The inset in Figure 8B⇓ shows the best-fit sigmoids for the CS-CaT relation in Ctl and P42 cells, with CS values normalized to the maximum CS values in each. The relations can be superimposed, and the bath Ca2+ concentration associated with 50% of maximum CS in Ctl cells (0.647 mmol/L) was very similar to that of P42 cells (0.644 mmol/L), suggesting that the intrinsic Ca2+ sensitivity of the contractile apparatus was not altered but the degree of CS at any given level of activator Ca2+ was reduced.
The present study shows that chronic rapid atrial activation, as occurs during AF, impairs the contractile performance of atrial myocytes. In addition, atrial tachycardia reduces the systolic Ca2+ transient in a frequency-related fashion. Similar results were obtained in cells from dogs with AF compared with those from dogs maintaining 1:1 pacing throughout the study period. The reduced Ca2+ transient contributes to contractile dysfunction but does not account for it completely.
Pathogenesis of Atrial Contractile Dysfunction
Although the delayed restoration of effective left atrial contraction after reversion of AF to sinus rhythm has been recognized for >30 years, its underlying mechanisms remain unclear. Contractile dysfunction has been thought to be related to intrinsic myocardial abnormalities2 3 4 6 and possibly to the mode of cardioversion (pharmacological versus electrical).12 23 Depressed atrial mechanical function after cessation of AF does, however, occur in patients without identifiable heart disease (lone AF),8 10 11 although its recovery may be faster in patients without associated structural heart disease.11 Furthermore, a recent study of AF patients and disease-matched control subjects has confirmed that transient atrial contractile dysfunction is due to AF per se.10 The impact of the mode of cardioversion on atrial mechanical function remains controversial23,24; however, a transient depression of atrial contractility has been well documented after both pharmacological and electrical conversion to sinus rhythm.9 10 11 23 24 Furthermore, transthoracic electrical cardioversion to terminate ventricular tachycardia does not impair atrial contractile function.25
The present study is the first of which we are aware to examine the changes in cellular contractility caused by sustained atrial tachycardia. Our results suggest that the rapid atrial activation rates in AF (generally 400 to 600 bpm) are sufficient to cause substantial impairments in atrial contractile function that in and of themselves may account for transient atrial contractile depression on reversion from AF to sinus rhythm. This idea is consistent with recent work that has shown that atrial flutter, which produces atrial rates of the order we studied, causes significant impairments in atrial contractile function.26 The contractile impairments we observed most likely explain the significant atrial dilation observed by Morillo et al16 in dogs subjected to 6 weeks of atrial pacing at 400 bpm. Short-term AF (15 minutes) has been shown by Leistad et al15 to reduce atrial contractility, a phenomenon antagonized by verapamil infusion. It is difficult to relate these findings for very brief AF to the slower-developing changes we noted. The latter are likely to involve downregulation of mRNA expression,27 whereas the shorter-term changes may be due to [Ca2+]i-induced ICa inhibition.28
Cellular Mechanisms Underlying Atrial Contractile Dysfunction Caused by Atrial Tachycardia
To gain insight into the cellular mechanisms of the atrial contractile dysfunction of paced dogs, we studied the cellular Ca2+ transient and the response of CS and the CaT to rate change. We found that the CaT is reduced in cells from paced dogs, almost certainly contributing to reduced contractility. This observation is compatible with previously described decreases in ICa of paced dogs.18 Our present observations, however, point to additional abnormalities in cellular contractility and Ca2+ handling. When CaTs were increased to normal values in P42 cells by increasing [Ca2+]o, CS failed to normalize, indicating an inability of the contractile apparatus to respond normally to the CaT. This abnormality is not due to an alteration in sensitivity to Ca2+ per se, because the normalized CS-CaT relations of P42 and Ctl cells can be superimposed (inset, Figure 8B⇑). The results are consistent with abnormal function of the contractile apparatus. The detailed mechanisms underlying contractile apparatus dysfunction, in terms of their biochemical and structural bases, remain to be determined. In addition to abnormalities in the amplitude of the CaT, paced cells show slowed kinetics of relaxation and markedly altered responses to repetitive activation. Some of these findings resemble those previously reported in ventricular failure29 and point to abnormal SR function. Important ultrastructural abnormalities have been noted by electron microscopy in the SR of rapidly paced dog atria,16 in keeping with our functional observations. More recent ultrastructural observations in goats with sustained AF indicate a loss of myofibrils, which may underlie the intrinsic contractile abnormalities we noted even when the CaT was normalized, along with fragmentation of the SR.30 The slowed relaxation of CS and the abnormal response of CS to rate increases closely parallel those of CaT, suggesting that these elements of contractile abnormality are most likely due to altered cellular Ca2+ handling.
We used the membrane-permeant acetoxymethyl ester form of indo 1 (indo 1-AM) to avoid the dialysis of intracellular contents that inevitably accompanies the loading of the free acid form via the pipette. Intracellular dialysis controls the cytoplasmic milieu and thus might mask possible alterations in cytoplasmic homeostasis and function caused by atrial tachycardia. In addition, cell dialysis can accelerate rundown of CS and the CaT, limiting the stability of measurements and the time intervals over which they can be meaningfully measured. Potential shortcomings of the AM form include the possibilities of intracellular compartmentation and partial deesterification of the indicator, which can alter cellular functions and limit the sensitivity of CaT measurement.20 We were careful to ensure adequate and equivalent loading of indo 1-AM to obtain accurate and reproducible measurements of CaTs. Although the indicator had mild effects on cellular contractility (Table⇑), our primary analyses of CS were performed on cells not exposed to indo 1, and the same changes in CS behavior were observed in paced cells whether or not they were loaded with the indicator.
It would have been interesting to know the absolute values of [Ca2+]i, rather than simply expressing [Ca2+]i in terms of the R400/500 ratio. It is possible to calibrate for [Ca2+] in aqueous solutions and then to apply this calibration to provide [Ca2+]i estimates; however, the properties of all Ca2+-sensitive dyes in the cytoplasm can be very different from those in aqueous solution.21 This can lead to considerable error in estimated [Ca2+]i values, and it has therefore become common practice to express the results of fluorescent Ca2+-sensitive dye studies in terms of fluorescent emission ratios (as we did), instead of as potentially misleading intracellular Ca2+ concentrations.
To evaluate the effects of rapid atrial activation on myofilament Ca2+ responsiveness, the CS-CaT relations were established for Ctl and P42 cells (Figures 7⇑ and 8⇑). We were unable to attain the plateau of the sigmoid relation curve, because exposing cells to higher extracellular Ca2+ concentrations caused frequent irregular spontaneous contractions. Nevertheless, the data were well fitted by classically described sigmoid relations22 that allowed for quantitative comparisons of the dependence of CS on CaT in each group of cells.
We have shown that sustained rapid atrial tachycardia, as occurs during AF, impairs cellular contractile function and Ca2+ handling. Some of the contractile dysfunction is due to an impaired systolic Ca2+ transient, but additional abnormalities of contractile apparatus function are also present. These findings provide the first detailed insights at the cellular level into mechanisms of the clinically relevant atrial contractile dysfunction associated with sustained AF.
Selected Abbreviations and Acronyms
|P42||=||dogs subjected to atrial pacing at 400 bpm for 42 days|
|P7||=||dogs subjected to atrial pacing at 400 bpm for 7 days|
|R400/500||=||ratio of emitted fluorescent light at 400/500 nm|
|ST dogs||=||dogs with 1:1 response to stimulation|
|TTP||=||time to peak|
This work was supported by the Medical Research Council of Canada, the Heart and Stroke Foundation of Quebec, and the Fonds de Recherche de l’Institut de Cardiologie de Montréal. Dr Leblanc is a Fonds de la Recherche en Santé du Québec Research Scholar, and Dr Gaspo is a Medical Research Council/Pharmaceutical Manufacturer’s Association of Canada Postdoctoral Fellow. The authors thank Emma De Blasio, Mirie Levi, and Nathalie Talbot for technical assistance, Antonio Guia for digital processing, and Caroll Boyer for secretarial help with the manuscript.
- Received October 2, 1997.
- Revision received January 6, 1998.
- Accepted March 12, 1998.
- Copyright © 1998 by American Heart Association
Graettinger JS, Carleton RA, Muenster JJ. Circulatory consequences of changes in cardiac rhythm produced in patients by transthoracic direct-current shock. J Clin Invest. 1964;43:2290–2302.
Ikram H, Nixon PGF, Arcan T. Left atrial function after electrical conversion to sinus rhythm. Br Heart J. 1968;30:80–83.
Lipkin DP, Frenneaux M, Stewart R, Joshi J, Lowe T, McKenna WJ. Delayed improvement in exercise capacity after cardioversion of atrial fibrillation to sinus rhythm. Br Heart J. 1988;59:572–577.
Grimm RA, Stewart WJ, Maloney JD, Cohen GI, Pearce GL, Salcedo EE, Klein AL. Impact of electrical cardioversion for atrial fibrillation on left atrial appendage function and spontaneous echo contrast: characterization by simultaneous transesophageal echocardiography. J Am Coll Cardiol. 1993;22:1359–1366.
Black IW, Fatkin D, Sagar KB, Khandheria BK, Leung DY, Galloway JM, Feneley MP, Walsh WF, Grimm RA, Stollberger C, Verhorst PMJ, Klein AL. Exclusion of atrial thrombus by transesophageal echocardiography does not preclude embolism after cardioversion of atrial fibrillation: a multicenter study. Circulation. 1994;89:2509–2513.
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.
Morillo CA, Klein GJ, Jones DL, Guiraudon CM. Chronic rapid atrial pacing: structural, functional, and electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation. 1995;91:1588–1595.
Yue L, Feng J, Gaspo R, Li GR, Wang Z, Nattel S. Ionic remodeling underlying action potential changes in a dog model of atrial fibrillation. Circ Res. 1997;81:512–525.
Fabiato A. Simulated calcium current can both cause calcium loading in and trigger calcium release from the sarcoplasmic reticulum of a skinned canine cardiac Purkinje cell. J Gen Physiol. 1985;85:291–320.
Spurgeon HA, Stern MD, Baartz G, Raffaeli S, Hansford RG, Talo A, Lakatta EG, Capogrossi MC. Simultaneous measurement of Ca2+, contraction, and potential in cardiac myocytes. Am J Physiol. 1990;258:H574–H586.
Abascal VM, Dubrey S, Ochoa MR, Davidoff R, Falk RH. Electrical vs pharmacologic cardioversion in atrial fibrillation: does the atrium really care? Circulation. 1995;92(suppl I):I-591. Abstract.
Dodds GA, Wilkinson WE, Greenfield RA, Natale A, Kisslo J, Pritchett E. Transthoracic cardioversion does not affect atrial mechanical function. Circulation. 1995;92(suppl I):I-591. Abstract.
Yue L, Wang Z, Gaspo R, Nattel S. The molecular mechanism of ionic remodeling of repolarization in a dog model of atrial fibrillation. Circulation. 1997;96(suppl I):I-180. Abstract.
Sun H, Leblanc N, Nattel S. Mechanisms of inactivation of L-type calcium channels in human atrial myocytes. Am J Physiol. 1997;272:H1625–H1635.
Ausma J, Wijffels M, Thoné F, Wouters L, Allessie M, Borgers M. Structural changes of atrial myocardium due to sustained atrial fibrillation in the goat. Circulation. 1997;96:3157–3163.