INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

LIFE-THREATENING VENTRICULAR ARRHYTHMIAS are common in human hearts with ischemic disease or heart failure composed of normal as well as damaged myocardium with variations of both excitation-contraction (E-C) coupling and local mechanical properties that coexist with electrical nonuniformity. Electrical nonuniformity has been extensively discussed as a substrate for these arrhythmias. Less is known, however, about the role played by nonuniformities of E-C coupling and myocardial mechanics in these arrhythmias (39).

We have shown that propagation of local aftercontractions [denoted as triggered propagated contractions (TPCs)], which can be measured as waves of the sarcomere shortening after a twitch, are observed in both rat ventricular (14, 34, 43) and human atrial trabeculae (13). TPCs propagate at a constant velocity along trabeculae with a regional increase in intracellular Ca²⁺ concentration ([Ca²⁺]_i), which have been denoted as Ca²⁺ waves. Such Ca²⁺ waves propagate much faster than those observed in single myocytes (33, 39), and the propagation velocity (V_prop) ranges from 0.34 to 5.47 mm/s (31) depending on both [Ca²⁺]_i and sarcoplasmic reticulum (SR) Ca²⁺ loading (32). Importantly, it has been reported that TPCs and Ca²⁺ waves cause delayed afterdepolarizations (DADs) and triggered arrhythmias and that the triggered arrhythmias occur often when TPCs propagate rapidly (39).

Ca²⁺ waves and TPCs invariably start from damaged regions within trabeculae, such as both ends of trabeculae or regions near cut side branches (39). The damaged cells and the border zone are weaker than those from the central regions of trabeculae during a twitch, that is, they exhibit nonuniform E-C coupling. Therefore, during a twitch, the central regions of the trabeculae stretch the damaged region, and, during relaxation, the damaged region shortens suddenly. Elimination of the stretch of the damaged region and border zone during the twitch and hence elimination of the quick release during the relaxation phase has been shown to prevent the induction of TPCs (14), suggesting that stretch or quick release of damaged ends of trabeculae during electrically stimulated twitch plays an important role in the triggering mechanism of TPCs. Therefore, we expect that the investigation of TPCs and Ca²⁺ waves will contribute to knowledge of the role of nonuniform E-C coupling and myocardial mechanics played in the triggering of ventricular arrhythmias in the diseased heart (12, 34, 39).

Stretch of cardiac muscle is known to cause an increase in force without a change in [Ca²⁺]_i, indicating that the increase in force is at least partially due to the length dependence of myofilament Ca²⁺ sensitivity (1, 27), although an initial rapid increase in force is followed by a slow increase with an increase in [Ca²⁺]_i over a period of minutes (3, 4, 8, 9). Quick reduction of the muscle length after a transient stretch can dissociate Ca²⁺ from the myofilament Ca²⁺-binding proteins due to the change in myofilament Ca²⁺ sensitivity (3, 23, 26, 29). We hypothesized that dissociation of Ca²⁺ especially occurs near the damaged ends within trabeculae and triggers TPCs, because the damaged cells and the border zone may be more easily stretched by a transient increase of the muscle length (39). On the other hand, it has been reported that an increase in [Ca²⁺]_i modulates the Ca²⁺-induced Ca²⁺ release (CICR) mechanism, which is known to be an important mechanism for the propagation of Ca²⁺ waves (6, 32). Therefore, we hypothesize that dissociated Ca²⁺ due to the reduction of myofilament Ca²⁺ sensitivity after a quick release may play a role in the initiation and propagation of TPCs and Ca²⁺ waves through the modulation of the CICR mechanism.

In the present study, focusing only on the rapid response to a transient change in muscle length, we first investigated the effect of a stretch (or release) pulse (~200 ms) during a twitch contraction on subsequent changes in [Ca²⁺]_i during TPCs to evaluate whether Ca²⁺ dissociated from myofilaments plays a role in the initiation and propagation of TPCs and Ca²⁺ waves. We also tested whether Gd³⁺ can prevent the changes in [Ca²⁺]_i caused by the transient stretch, because Gd³⁺ has been reported to block mechanosensitive membrane channels [stretch-activated channels (SACs)] (24, 36, 40-42), which may conceivably play a role in the observed responses in this study.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Dissection and mounting of rat ventricular trabeculae. The experiments were performed according to the guidelines for the care and use of laboratory animals of Tohoku University. Sprague-Dawley rats (250-300 g) were anesthetized with an intraperitoneal injection of pentobarbital sodium (40 mg/kg). Hearts (n = 15) were excised, and the coronary arteries were immediately perfused via the aorta with a HEPES-buffered solution containing (in mM) 130.8 NaCl, 5 KCl, 1.2 MgCl₂, 1.2 Na₂SO₄, 2.8 acetate, 10 HEPES, 10 glucose, and 0.7 CaCl₂. The solutions were bubbled with 100% O₂; pH was adjusted to 7.4 by addition of NaOH (42). After the heart was arrested by flushing with the same solution modified by addition of 15 mM KCl, trabeculae (n = 15, slack length, 1.45 ± 0.07 mm; slack width, 334 ± 27 µm; slack thickness, 114 ± 6 µm) were dissected from the right ventricle and mounted horizontally between a force transducer and a micromanipulator with a direct current torque motor (type 3123, No. 8-687, SAN-EI Instruments) in a perfusion bath located on the stage of an inverted microscope (Nikon) (31, 32). The trabeculae were equilibrated at 0.5 Hz with stimuli through parallel platinum electrodes in the bath with 5-ms pulses 50% above the threshold and superfused with HEPES-buffered solution [extracellular Ca²⁺ concentration ([Ca²⁺]_o) = 0.7 mM] at room temperature (23-28°C).

Fura 2 loading and measurement of fluorescence. [Ca²⁺]_i was measured as previously described (7, 31, 32). Briefly, fura 2 pentapotassium salt was microinjected electrophoretically into one cell and allowed to spread throughout the trabeculae via gap junctions. After the injection, the trabeculae were stimulated at 1 Hz for 30-60 min until the fura 2 had diffused uniformly throughout the preparation. The epifluorescence of fura 2 from the trabecula at excitation wavelengths of 340 and 380 nm was measured at 510 nm by a photomultiplier tube (PMT; E1341 with a C1556 socket, Hamamatsu). The signal from the PMT was stored (RD-130TE DAT Data Recorder, TEAC) for later analysis. Alternatively, the fluorescent image of the trabecula at excitation wavelengths of 360 and 380 nm was recorded by a SIT video camera (Hamamatsu C2400-8, Hamamatsu) through a 510-nm band-pass filter (25, 33). The images were recorded with a videocassette recorder (VCR) for off-line analysis.

Length changes and force measurements. Force was measured using a silicon semiconductor strain gauge and expressed as stress after normalization for the cross-sectional area of the muscle measured in slack conditions. Sarcomere length (SL) was measured using laser diffraction techniques (38). Changes in muscle length were introduced at the valvular end of trabeculae using a direct current torque motor driven by voltage commands from a personal computer via a 12-bit digital-to-analog converter (T-IFPC9 04120A-N, TSK). In this study, stretches were adjusted in the unstimulated muscle to cause 5 or 10% changes of SL. Thus we defined the length changes from 2.0 to 2.1 or 2.2 µm of resting SL as 5 or 10% stretches, respectively. After 10% stretch, peak active force increased by 68.0 ± 10.1% during electrically stimulated twitches at a 2-s stimulation interval ([Ca²⁺]_o = 2.0 mM). Passive force was 2.5 ± 0.3% of the peak total force at a SL of 2.0 µm (no stretch) and 8.4 ± 1.0% of the peak total force at a SL of 2.2 µm (10% stretch). "Active force" was defined as total force (stress measured during a contraction) minus passive force (stress measured at a resting condition). In addition to the stretch protocol described above, the muscle was shortened by 10% (10% release). In this release protocol, trabeculae were released from 2.0 µm to an almost slack SL (~1.8 µm).

Membrane potential measurements. Membrane potential was measured using ultracompliant glass microelectrodes as described elsewhere (12, 17). Flexible stepped electrodes drawn with a glass microelectrode puller (Narishige) were filled with 3 M KCl and connected to a high-input impedance amplifier (MEZ-8201, Nihon Koden). The center of the trabecula was impaled. Long-term stable measurements were possible using these electrodes without causing local damage either during contraction or during changes of muscle length.

Analysis of the signal from the PMT. [Ca²⁺]_i was determined using the following equation (19), as previously described (31, 32)

(1)

where K_d is the effective dissociation constant, R is the ratio of the fluorescence (after subtraction of the autofluorescence of the muscle) at 340-nm excitation to that at 380-nm excitation, R_min is R at zero [Ca²⁺], R_max is R at a saturating [Ca²⁺], and  is the ratio of the fluorescence value for Ca²⁺-free dye to the fluorescence value for Ca²⁺-bound dye at 380-nm excitation. Values for K_d, R_min, R_max, and  were determined using in vitro calibrations, because we previously reported a good correlation between in vitro and in vivo calibrations when the free Mg²⁺ was 1 mM in the solutions mimicking the intracellular milieu (7). R_min was 0.48, R_max was 5.98, K_d was 382 nM, and  was 5.23. This K_d value is in agreement with previous data (31).

Analysis of fura 2 images. Fura 2 fluorescence images recorded at 30 frames/s on the VCR were analyzed as previously described (31, 32). Briefly, the fluorescence data of each video frame were digitized with an 8-bit analog-to-digital converter and stored in a frame buffer memory of 512 × 512 pixels (FRM1-512 Image Digitizer, Photron). Therefore, in our optical system, one pixel corresponded to 2.05 × 2.05 µm in the image plane. For analysis of the image, a region of interest (ROI) was set horizontally along the long axis of the fluorescence image of the trabecula. The length of the ROI was always 512 pixels (1,049 µm), while its width was 20 pixels (41 µm). To obtain intensity profiles of the fluorescence along the long axis of the trabecula, we calculated the average intensity value from each transverse line of pixels within the ROI. To eliminate high-frequency noise from the intensity profile, we used a low-pass finite impulse response filter (MATLAB) with a cutoff frequency of 5 pixels (10 µm). After the autofluorescence was subtracted, we calculated the ratio of the fluorescence at 360-nm excitation to that at 380-nm excitation at each point on the intensity profiles obtained from the images at 360- and 380-nm excitation light. To correct for the effects of nonuniform illumination of excitation light, we calculated [Ca²⁺]_i at each sampling point after the induction of TPCs using the regression line derived from the relationship between the PMT and the SIT ratio determined in the same sampling region. Furthermore, to avoid noise caused by low-excitation light intensity on the far edges of the profile of fluorescence, we calculated [Ca²⁺]_i only for the centrally located 300 pixels (615 µm) of the ROI. Thus we analyzed regional [Ca²⁺]_i in the center of trabeculae where the muscles were intact and the propagation of Ca²⁺ waves along trabeculae was observed.

Induction of TPCs. The resting SL was set at 2.0 µm and the temperature at 28°C before the induction of TPCs. To induce TPCs, trains of electrical stimuli at an interval of 500 ms were applied for 7.5 s at [Ca²⁺]_o = 2.0 mM. This stimulus protocol was repeated every 15 s. The stimulus interval was then shortened in steps of 20-30 ms until a TPC appeared (14, 34). The measurement of [Ca²⁺]_i started when the V_prop of TPCs triggered by serial trains of stimuli varied by <10%. Reproducible conditions lasted at least 30 min (31, 32). In this study, we induced reproducible TPCs in 15 trabeculae from 15 hearts. To change the Ca²⁺ loading of the muscle, we further shortened the stimulus interval of the train of electrical stimuli in steps of 20-30 ms in five trabeculae. The measurement of [Ca²⁺]_i started when TPCs had reached a steady state. Thus we eventually analyzed 27 reproducible TPCs induced by trains of electrical stimuli at intervals of 378 ± 19 ms (temperature, 27.9 ± 0.2°C).

Experimental protocol. First, when reproducible TPCs were observed, the trabecula was stretched transiently by 5 or 10% or released by 10% (n = 13). The stretch (or release) pulses were applied 10 ms after the last stimulus of the electrical train. The duration of stretch (or release) pulses varied from 150 to 200 ms (see Fig. 1A), depending on the duration of the last stimulated twitch, to adjust the moment of release to the end of the relaxation phase (corresponding to 80-90% relaxation of active force) when its effect was maximal. Second, trabeculae were also stretched transiently by 5 or 10% during TPCs. In this case, the stretch pulses were applied 300 ms after the last stimulus of electrical trains for 500 ms (see Fig. 6A) to study the direct effect of the stretch on [Ca²⁺]_i underlying the TPC (n = 9). Third, to investigate the role of membrane stretch of this magnitude on SACs, we applied the stretch pulse during the last twitch in the presence of 10 µM Gd³⁺, the most widely used SAC inhibitor (24, 36, 40, 41). After the superfusion with 10 µM Gd³⁺ at [Ca²⁺]_o = 2.0 mM, force development induced by trains of electrical stimuli decreased and Ca²⁺ waves disappeared (42). To induce TPCs again, [Ca²⁺]_o was increased in steps of 1 mM until TPCs reappeared. We finally induced five reproducible TPCs in five trabeculae at [Ca²⁺]_o = 5.2 ± 0.73 mM (temperature, 27.6 ± 0.5°C). Under these conditions, active force and peak [Ca²⁺]_i by the last twitch of the series were restored to 83.8 ± 9.1 and 93.5 ± 7.4% of the level before superfusion with Gd³⁺, respectively.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Effect of transient changes in the muscle length during the last twitch of electrical trains on force and intracellular Ca2+ concentration ([Ca2+]i). Red line, 10% stretch; green line, 5% stretch; black line, no stretch; blue line, 10% release. A, top: force recordings (normalized for cross-sectional area of the muscle) during the last two electrically stimulated twitches (stimulation interval of 280 ms) and a subsequent triggered propagation contraction (TPC); middle, changes in [Ca2+]i calculated from the fluorescence signals recorded by the photomultiplier tube (PMT); bottom, changes in muscle length. As shown in top, with 5 and 10% transient stretch pulses for 150 ms, the peak TPC (PeakTPC) increased from 5.8 to 6.3 and 7.4 mN/mm2, respectively, whereas the time between the last stimulus and the PeakTPC (TimeTPC) decreased from 298 to 280 and 274 ms, respectively. In contrast, with a 10% release pulse for 150 ms, PeakTPC decreased to 5.4 mN/mm2, whereas TimeTPC increased to 374 ms. As shown in middle, with 5 and 10% transient stretch pulses for 150 ms, PeakCW increased from 296 to 330 and 391 nM, respectively, whereas TimeCW decreased from 278 to 260 and 244 ms, respectively. With a 10% release pulse for 150 ms, PeakCW decreased to 238 nM, whereas TimeCW increased to 326 ms. FT, active force by the last stimulus of the electrical trains; PeakCW, peak of the increase in [Ca2+]i during TPCs; TimeCW, the time between the last stimulus and PeakCW; S, moment of electrical stimulation (indicated by arrow). B: expansion of A, middle. Red arrow, the quick release (QR) transient ("QR-transient"). Temperature, 28.3°C; Exp. 991101.

Data analysis. To assess the latency of development of Ca²⁺ waves or TPCs, we calculated the following parameters (see Figs. 1 and 5A). First, we measured the peak of the TPC force (Peak_TPC) and the time between the last stimulus and the Peak_TPC (Time_TPC). Second, we measured the peak of the increase in [Ca²⁺]_i during TPCs (Peak_CW) and the time between the last stimulus and the Peak_CW (Time_CW). Third, we measured the peak of the DADs (Peak_DAD) and the time between the last stimulus and the Peak_DAD (Time_DAD). We also calculated the active force of the last twitch of the train (F_T) and used the changes in F_T by the transient stretch as an indicator of the Ca²⁺ bound to the myofilament Ca²⁺-binding proteins (2, 35).

Statistics. All measurements are expressed as means ± SE. Statistical analysis was performed using ANOVA and paired Student t-tests as appropriate. Values of P < 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Figures 1 and 2 show that a transient change in muscle length during the last electrically stimulated twitch of the train can affect the latency of the TPC force and Ca²⁺ waves. Figure 1 shows an example of force development and changes in the [Ca²⁺]_i during the last two electrically stimulated twitches of a train at a stimulation interval of 280 ms. It is clear that the penultimate twitch was followed by a slow and small diastolic increase in force on which the last twitch was superimposed. The last twitch by itself was followed by a force transient with a time course similar to those that have been reported in the past as force transients accompanying TPCs (14, 34, 39). Indeed, we observed the occurrence of propagated sarcomere shortening as well as propagating [Ca²⁺]_i waves (data not shown) as have been shown previously during TPCs (31, 32, 39). It was clear that TPCs could be elicited at 28°C by a suitable combination of stimulus conditions (stimulus rate, train length, and [Ca²⁺]_o) as chosen in this study. This confirms that TPCs can occur at a wide range of temperatures. The time over which the TPCs behave in a stable fashion decreased with increasing temperature (15). We observed in these experiments that the stable period for TPCs exceeded 30 min.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Effect of a transient stretch of the muscle during the last electrically stimulated twitch on regional changes in [Ca2+]i. The regional [Ca2+]i was calculated a few minutes after the recording using the PMT shown in Fig. 1. x-Axis, time; y-axis, [Ca2+]i; z-axis, position along the long axis of the trabecula; S, moment of electrical stimulation (indicated by arrow). A, top: muscle length; bottom, three-dimensional presentation showing the regional changes in [Ca2+]i in the absence of the stretch pulse. After the end of the clearly uniform stimulated Ca2+ transient, a small Ca2+ transient was observed to move as a wave at the calculated propagation velocity (Vprop) of 2.9 mm/s. B, top: muscle length; bottom, three-dimensional presentation showing the regional changes in [Ca2+]i in the presence of a 10% stretch pulse for 150 ms during the last electrically stimulated twitch. Vprop increased to 6.77 mm/s. Exp. 991101.

The time course of the TPCs appeared to reflect the time course of an underlying [Ca²⁺]_i transient (Fig. 1), as has been observed previously (31, 32). With a stretch pulse by 5 or 10% for 150 ms during the last of the train, force increased, whereas Ca²⁺ transients during the twitch showed no changes in their peaks but exhibited a faster decline and lower minimal [Ca²⁺]_i. With a release pulse for 150 ms, the force of the twitch decreased, and Ca²⁺ transients during the twitch exhibited a slower decline with no changes in their peaks. The stretch pulse accelerated and the release pulse suppressed the subsequent TPC force and underlying [Ca²⁺]_i. Here, it should be noted that the acute return of muscle length, especially after the 10% transient stretch, induced a small and rapid transient increase (denoted as "QR transient" in Fig. 1B) in [Ca²⁺]_i, followed by a larger and earlier increase in [Ca²⁺]_i during the TPC.

To study the details of changes in the TPC force and the underlying increase in [Ca²⁺]_i, spatial changes in [Ca²⁺]_i during comparable TPCs were determined directly after the recordings (such those shown in Fig. 1). Figure 2 shows spatial changes in [Ca²⁺]_i during the last electrically stimulated twitches of a train at a stimulation interval of 280 ms and a Ca²⁺ wave. With transient stretch by 10% during the last twitch, V_prop of the Ca²⁺ wave increased from 2.90 to 6.77 mm/s. With transient stretch by 5% for 150 ms, V_prop of the Ca²⁺ wave increased from 2.90 to 4.29 mm/s (data not shown). We could not measure V_prop of the Ca²⁺ wave after a release pulse, because the Ca²⁺ wave was too small to detect the peak of [Ca²⁺]_i at each position along the trabeculae.

In Fig. 3, the open circles show the effect of changes in the muscle length during the last twitch of the trains on TPC force and the underlying changes in [Ca²⁺]_i obtained from nine trabeculae (n = 13). Transient stretch decreased Time_TPC and Time_CW and increased Peak_TPC and Peak_CW, whereas the transient release increased Time_TPC and Time_CW and decreased Peak_TPC and Peak_CW. These observations show that transient stretch can accelerate and the release pulse can suppress the subsequent TPC force and the underlying increase in [Ca²⁺]_i.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Effect of changes in muscle length during the last electrically stimulated twitch on the latency of TPC force and the underlying increase in [Ca2+]i obtained from 13 TPCs from 9 trabeculae. , Changes in the absence of Gd3+; , changes in the presence of 10 µM Gd3+. With a 5 or 10% stretch pulse, TimeTPC (P < 0.001 or P < 0.0005, respectively) and TimeCW (P < 0.0005 or P < 0.0001, respectively) decreased, whereas PeakTPC (P < 0.05 or P < 0.05, respectively) and PeakCW (P < 0.005 or P < 0.005, respectively) increased. With the 10% release pulse, TimeTPC (P < 0.001) and TimeCW (P < 0.001) increased, whereas PeakTPC (P < 0.01) and PeakCW (P < 0.005) decreased. In the presence of 10 µM Gd3+, TimeTPC (P < 0.05) and TimeCW (P < 0.05) decreased significantly, whereas PeakTPC (P < 0.05) and PeakCW (P < 0.01) increased significantly with the 10% stretch pulse. The changes in the presence of Gd3+ were not significantly different from those in the absence of Gd3+.

Figure 4 shows the changes of Ca²⁺ waves owing to transient stretch of the muscle during the last twitch of the trains. In Fig. 4A, we selected the peak of the Ca²⁺ transient at each pixel along the trabecula during the Ca²⁺ waves illustrated in Fig. 2 and superimposed the plots of the peak time against the peak position obtained from both Ca²⁺ waves to investigate the changes in the propagation of Ca²⁺ waves owing to the transient stretch. We noticed that, after the stretch pulse, Ca²⁺ waves started from closer to the end of the trabecula and propagated faster along the trabecula at a constant velocity. Figure 4B shows the effect of transient stretch by 5 or 10% on V_prop of nine Ca²⁺ waves obtained from five trabeculae. The V_prop of Ca²⁺ waves increased significantly by the transient stretch. When we plotted the changes in V_prop of Ca²⁺ waves against those in F_T by comparable stretch (Fig. 4C), it appeared that the changes in V_prop correlated strongly (r = 0.84) with those in F_T, i.e., changes in the amount of Ca²⁺ bound to the myofilament Ca²⁺-binding proteins.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effect of a transient stretch during the last electrically stimulated twitch on Ca2+ waves. A: relationships between the peak time and peak position of the Ca2+ transient along the trabecula obtained from Ca2+ waves in Fig. 2.  and , Relationships obtained from the Ca2+ waves without () and with the transient 10% stretch (). Relationships were fitted by linear regression; the slope of the fitted line was taken as the velocity of Ca2+ waves. Dotted vertical line, moment when the muscle was released after the transient stretch. B: effect of a transient stretch during the last electrically stimulated twitch on Vprop of 9 Ca2+ waves obtained from 5 trabeculae. Vprop of Ca2+ waves increased significantly after the 5 or 10% stretch. C: relationship between changes in Vprop of Ca2+ waves and those in FT caused by the transient stretch during the last electrically stimulated twitch (n = 11). Changes in Vprop linearly correlated with those in FT (r = 0.84, P < 0.01).

Figure 5A shows the effect of the stretch pulse during the last twitch of the trains on the membrane potentials. The resting membrane potential of this muscle was ~70 mV before the stimulus train, was depolarized during the stimulus train (~50 mV), and then gradually returned to the basal level. We observed DADs after the cessation of electrical stimuli concurrent with the TPC as has been shown previously (12). With a transient stretch of muscle by 10% for 200 ms, the DAD occurred earlier and appeared larger. Data (n = 9) obtained from three trabeculae showed that the resting membrane potential before the stimulus train was 69.7 ± 3.5 mV and that, with the transient stretch, Time_DAD decreased by 7.0 ± 1.9% and Peak_DAD increased from 55.7 ± 3.3 to 54.7 ± 3.3 mV (P < 0.0005).

View larger version (32K): [in this window] [in a new window]   Fig. 5.   A: effect of a transient 10% stretch of the muscle for 200 ms during the last twitch of electrical trains on force and delayed afterdepolarizations (DADs). Thick line, 10% stretch; thin line, no stretch. Top, force; top middle and bottom middle, changes in membrane potential and the expansion of the recording in the ordinate recorded by an ultracompliant glass microelectrode during the last two electrically stimulated twitches (stimulation interval of 300 ms) and subsequent TPC; bottom, muscle length. PeakDAD, peak DAD; TimeDAD, time from the last stimulus to PeakDAD. TimeDAD decreased from 350 to 298 ms, and PeakDAD increased 58.9 to 57.7 mV as a result of transient 10% stretch (thick line) of the muscle. S, moment of electrical stimulation (indicated by arrow). Temperature, 27.4°C; Exp. 000511. B: effect of transient changes in the muscle length for 300 ms during the last twitch of electrical trains (stimulation interval of 350 ms) on force and [Ca2+]i in the presence of 1 mM caffeine. Top, top middle, bottom middle, and bottom are the same as those shown in A. Red line, 10% stretch; black line, no stretch; blue line, 10% release. Red arrow, increase in [Ca2+]i at the moment of quick return of muscle length after the stretch pulse. Temperature, 28.1°C; Exp. 990902.

To eliminate TPCs and Ca²⁺ waves and clarify the Ca²⁺ dissociation from the myofilaments, we added the stretch or release pulse during the last stimulated twitch in the presence of 1 mM caffeine (Fig. 5B). In the presence of 1 mM caffeine, the amplitude of stimulated twitch and Ca²⁺ transient decreased, and TPCs and Ca²⁺ waves disappeared as has been observed previously (32). With the stretch (or release) pulse, we can observe a small and quick transient increase (or decrease) in [Ca²⁺]_i at the quick return of the muscle length after 10% transient stretch (or release), respectively. In three trabeculae, we changed the muscle length during the last stimulated twitch in the presence of 1 mM caffeine and observed the same transient increase (or decrease) in [Ca²⁺]_i at the quick return of the muscle length. We therefore assumed that the small and transient increase in [Ca²⁺]_i was caused by Ca²⁺ dissociation from the myofilaments and occurs as the QR transient observed in the absence of caffeine.

It has been reported that V_prop of Ca²⁺ waves is determined by both the open probability of the SR Ca²⁺ release channels and cellular Ca²⁺ loading (32). To study the direct effect of the transient stretch pulse on the open probability of SR Ca²⁺ release channels, we stretched the muscle length during a TPC (Fig. 6A). The transient stretch during the TPC changed neither Time_TPC, Time_CW, nor Peak_CW but increased Peak_TPC. Data obtained from nine trabeculae showed that transient stretch of the muscle during TPCs accelerated neither the TPCs nor the underlying increase in [Ca²⁺]_i (data not shown).

View larger version (32K): [in this window] [in a new window]   Fig. 6.   A: effect of a transient 10% stretch of the muscle during TPCs on force and [Ca2+]i. Thick line, 10% stretch; thin line, no stretch. Top, force; middle, [Ca2+]i; bottom, muscle length during the last two electrically stimulated twitches (stimulation interval of 400 ms) and subsequent TPC. TimeTPC, TimeCW, and PeakCW showed no changes, whereas PeakTPC increased as a result of the transient 10% stretch (thick line) for 500 ms during TPCs. S, moment of electrical stimulation (indicated by arrow). Temperature, 28.4°C; Exp. 990531. B: effect of a transient stretch of the muscle during the last twitch of electrical trains on force and [Ca2+]i in the presence of 10 µM Gd3+. Thick line, 10% stretch; thin line, 5% stretch; intermediate line, no stretch. Top, force; middle, [Ca2+]i; bottom, muscle length during the last two electrically stimulated twitches (stimulation interval of 320 ms) and a subsequent TPC. TimeTPC and TimeCW shortened, whereas PeakTPC and PeakCW increased as a result of transient stretch of the muscle for 180 ms. Inset, arrowhead showing the small QR transient of [Ca2+]i. Temperature, 27.7°C; extracellular Ca2+ concentration, 5.0 mM; Exp. 990708.

To investigate whether SACs play a role in the acceleration of TPC forces and the underlying increase in [Ca²⁺]_i, we used Gd³⁺, as shown in Figs. 3 and 6B. Figure 6B shows an example of force developments and changes in [Ca²⁺]_i during the last two electrically stimulated twitches of a train at a stimulation interval of 320 ms in the presence of 10 µM Gd³⁺. With a transient stretch of muscle by 5 or 10% for 180 ms during the last stimulated twitch of the train, the developed force and Ca²⁺ transients during the last twitch were affected to the same extent as those in the absence of Gd³⁺. In addition, proceeding from a small QR transient of [Ca²⁺]_i (see above), the acceleration of the TPC force and the underlying increase in [Ca²⁺]_i owing to the stretch were also observed. In Fig. 3, the open squares show the effect of the transient stretch on TPC forces and the underlying increase in [Ca²⁺]_i from five trabeculae in the presence of 10 µM Gd³⁺. After the rapid stretch, Time_TPC and Time_CW still decreased, and Peak_TPC and Peak_CW increased (n = 5). In addition, these changes in the presence of Gd³⁺ were not significantly different from those in the absence of Gd³⁺. Thus 10 µM Gd³⁺ did not suppress the acceleration of the TPC force and the underlying increase in [Ca²⁺]_i induced by the transient stretch.

Figure 7 shows that the transient length changes can be arrhythmogenic (18, 21). The TPC force and the underlying increase in [Ca²⁺]_i were accelerated by the transient stretch pulse during the last stimulated twitch of trains with stimulation intervals of 350 ms. It was striking that, after the transient stretch of the muscle length by 10% for 200 ms, a Ca²⁺ transient and force transient similar to those induced by electrical stimulation occurred without electrical stimuli, preceded by enhanced TPC force and an increase of the [Ca²⁺]_i transient. In addition, V_prop of the Ca²⁺ wave calculated in this trabecula was 8.82 mm/s without stretch and increased to 9.74 mm/s after a 5% transient stretch. With the 10% stretch pulse, the Ca²⁺ wave propagated too fast to be calculated from video frames obtained at 30 frames/s. Moreover, [Ca²⁺]_i showed synchronous changes throughout the trabecula during the extra twitch (data not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Effect of a transient stretch of the muscle on the induction of arrhythmias. Thick line, 10% stretch; thin line, 5% stretch; intermediate line, no stretch. Top, force; middle, [Ca2+]i; bottom, muscle length during the last two electrically stimulated twitches (stimulation interval of 350 ms) and subsequent TPC. Note that a twitch contraction was induced (arrowheads) as a result of transient 10% stretch of the muscle for 200 ms. S, moment of electrical stimulation (indicated by arrow). Temperature, 28.1°C; Exp. 991004.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This is the first study to characterize the effect of Ca²⁺ dissociated from myofilaments into the cytosol on TPCs and Ca²⁺ waves by direct measurement of the spatial and temporal properties of Ca²⁺ changes in a cardiac multicellular preparation. Our observations suggest that 1) a transient stretch (or release) pulse of the muscle length during the last twitch of electrical trains can accelerate (or suppress) the TPC force and the underlying increases in [Ca²⁺]_i through changes in V_prop of Ca²⁺ waves and 2) the changes in V_prop of Ca²⁺ waves can be explained by Ca²⁺ dissociated from myofilaments into the cytosol after rapid shortening but by neither a direct increase in the open probability of the SR Ca²⁺ release channel, SACs, nor voltage-gated channels, as will be discussed below.

Effect of stretches on stimulated twitches. It has been observed that a rapid response in force to a change in the muscle length occurs without a change in [Ca²⁺]_i using rat ventricular muscles (3, 5, 27), cat papillary muscles (3), and ferret ventricular muscles (29), and it has been reported that most of the change in force by stretch is due to the length dependence of myofilament Ca²⁺ sensitivity (1, 3, 27). Thus the change in twitch force during a stretch (or release) pulse of the muscle length observed in this study (Fig. 1) was at least partially due to the change in myofilament Ca²⁺ sensitivity. The faster (or slower) decline rate of Ca²⁺ transients during the stretch (or release) pulse was also due to the change in myofilament Ca²⁺ sensitivity (11, 28, 30).

Effect of transient stretch (or release) pulses on TPCs and Ca²+ waves. When Ca²⁺-activated cardiac muscle is allowed to shorten, the amount of Ca²⁺ dissociated from myofilaments depends on the changes in force or the number of force-generating cross-bridges rather than on changes in muscle length (2, 35). Dissociated Ca²⁺ may change cellular Ca²⁺ loading, which is one of the determinants of V_prop of Ca²⁺ waves (6, 14, 32). Thus the observation that the changes in the active force (F_T) by the transient stretch pulse correlated strongly with those in V_prop (Fig. 4C) suggests that the amount of Ca²⁺ dissociated from myofilaments can determine V_prop of Ca²⁺ waves.

No effect of SACs on the acceleration of Ca²+ waves. Several reports (24, 36, 41) have suggested that SACs are active in the heart because Gd³⁺ has been found to block stretch-induced phenomena in many cardiac systems. SACs (20, 24, 41) in the heart are cation selective, and the majority of them are permeable to Ca²⁺ (24, 36). In this study, however, SACs are unlikely to play an important role in the increase of [Ca²⁺]_i through the muscle for the following reasons: 1) the stimulated Ca²⁺ transient (Fig. 1) as well as [Ca²⁺]_i during TPCs (Fig. 6A) showed no increase during a stretch pulse; 2) the duration of the action potential did not become longer by the stretch pulse (Fig. 5A); and 3) 10 µM Gd³⁺ did not block the acceleration of the TPC force and the underlying changes in [Ca²⁺]_i induced by the transient stretch of the muscle (Figs. 3 and 6B). The latter result was consistent with our previous observation that Gd³⁺ sensitive SACs are involved in neither the triggering nor propagation processes of TPCs (42). In this study, we tested the full Gd³⁺ dose-response curve and observed that the effect of Gd³⁺ at 10 µM was 90% of the maximal effect. However, the effect of lowering [Ca²⁺]_o was identical to that of Gd³⁺ at all concentrations used, suggesting the absence of a specific effect due to SAC inhibition (42). In the present series of experiments, we corrected for the decrease of twitch force and the corresponding [Ca²⁺]_i transient by raising [Ca²⁺]_o to 5.2 mM to compensate (to within 15%) for the nonspecific inhibitory effect of Gd³⁺, especially on L-type Ca²⁺currents.

Other possible mechanisms for the acceleration of Ca²+ waves. First, stretch of the muscle may directly modulate the open probability of SR Ca²⁺ release channels, because it has been suggested that decreasing the SL below optimum (2.2 µm) results in a partial inhibition of CICR from the SR in cardiac muscle (16, 35). The transient stretch during the TPC, however, did not change [Ca²⁺]_i during TPCs (Fig. 6A), suggesting that the stretch pulse cannot modulate directly the open probability of SR Ca²⁺ release channels.

TPCs, DADs, and clinical relevance. Figure 5A confirms that TPCs are accompanied by DADs (12). It has been reported that the increase in [Ca²⁺]_i underlying a TPC is accompanied by a DAD via electrogenic Na⁺/Ca²⁺ exchange and activation of Ca²⁺-sensitive nonselective channels in the sarcolemma (37, 39). The amplitude of the DAD has been shown to correlate tightly with the amplitude of TPC. In turn, DADs may trigger action potentials (12, 37). Thus the rapid increase in [Ca²⁺]_i upon the reduction of the muscle length after the stretch may accelerate TPCs and the underlying increase in [Ca²⁺]_i and, thereby, induce arrhythmias, as we have shown here (Fig. 7).

Limitations of the study. We observed a small QR transient in [Ca²⁺]_i using PMT, which may have been caused by Ca²⁺ dissociated from myofilaments into the cytosol, whereas we could not find comparable changes in Ca²⁺ waves using the SIT camera. This difference between the recordings may have occurred for the following reasons. First, the field of view using the PMT is larger (diameter of ~1.3 mm) than that using the SIT camera (see MATERIALS AND METHODS) such that the initiating event of Ca²⁺ waves, which occurred usually at the ends of the trabecula, was out of view with the SIT camera but "visible" to the PMT. Second, for the analysis of the image, we defined the ROI (615 × 41 µm) horizontally along the long axis of the trabecula. If the initiating events of Ca²⁺ waves occurred outside of the ROI, we would not be able to detect these events. Third, the rising phase of the small QR transient of [Ca²⁺]_i lasted for ~20 ms, which was too short to be detected from video frames obtained at 30 frames/s. Thus, to observe directly how Ca²⁺ dissociated from myofilaments into the cytosol modulates Ca²⁺ waves, we must await exploration using probes with a higher spatial and temporal resolution.