All experiments were performed according to institutional guidelines. Male cardiomyopathic hamsters (Bio14.6) and normal F1B hamsters purchased from Bio Breeders Inc. (Fitchburg, MA, USA) and aged 36 to 57 weeks, were used for the experiments. Hamsters were anesthetized with intraperitoneal injection of sodium pentobarbital (50 mg/kg). A mid-line thoracotomy was quickly performed along with removal of heart and lungs. The PVs were perfused in a retrograde manner via polyethylene tube passed through the aorta and left ventricle into the left atrium. The proximal end of the polyethylene tubing was connected to a Langendorff apparatus for perfusion with oxygenated Tyrode's solution at 37°C. The Tyrode's solution contained (in mmol/L) NaCl 137, KCl 5.4, CaCl₂ 1.8, MgCl₂ 0.5, HEPES 10 and glucose 11 (pH was adjusted to 7.4 with NaOH) for about 15 minutes until efferent fluid was without blood.

The perfusate was then replaced with oxygenated Ca²⁺-free Tyrode's solution containing 1 mg/ml collagenase (Sigma, Type I) and 0.01 mg/ml protease (Sigma, Type XIV) for 30-40 minutes. Afterwards, the heart was washed with oxygenated Ca²⁺-free Tyrode's solution for 10 minutes. After that, the LA-PV area was removed from the heart, cut into fine pieces and gently shaken in 5-10 ml high-K⁺ storage solution until single cardiomyocytes were obtained.

Only LA-PV cardiomyocytes with clear cross striations obtained from 19 myopathic hamsters and 22 healthy hamsters were used for electrophysiological studies. Action potentials and ionic currents were recorded by means of whole-cell patch-clamp techniques with an Axopatch 1D amplifier (Axon Instruments, CA, U.S.A) as described in detail recently 11 13 15 . The standard pipette solution contained (in mmol/L) KCl 20, K aspartate 110, MgCl₂ 1, EGTA 0.5, Mg₂ATP 5, Na₂phosphocreatine 5, LiGTP 0.1, and HEPES 10, adjusted to pH 7.2 with KOH. The standard extracellular solution was the normal Tyrode's solution also used in cell isolation. Ionic currents were recorded in voltage-clamp mode. For the recording of the calcium currents I_Ca,L and I_Ca,T, tetraethylammonium (TEA) chloride and CsCl replaced NaCl and KCl, respectively. For I_Ca,T recording, tetrodotoxin (5 μmol/L) was added to block the fast Na⁺ current (I_Na). For K⁺ current measurement, CdCl₂ (200 μmol/L) was added to block I_Ca,L. A 30-ms prepulse from -80 to -40 mV was used to inactivate the sodium channel, followed by a 300-ms test pulses in 10-mV increments to +60 mV. I_K1 was quantified as 1 mmol/L Ba²⁺-sensitive current.

Action potentials (APs) were recorded in current-clamp mode. The tip potentials were zeroed before formation of the membrane-pipette seal in Tyrode's solution. After rupture, junction potential (8 mV) was corrected for AP recording. A small hyperpolarizing step from a holding potential of -50 mV to a testing potential of -55 mV was used to obtaining the total cell capacitance at the beginning of each experiment. The area under the capacitative current was divided by the applied voltage step to obtain the total cell capacitance. Normally 60-80% of series resistance (R_S) was electronically compensated. After compensation, the average time constant was 71 ± 6 μs, (cell capacitance 33 ± 2 pF, for 83 cells in healthy group; and 47 ± 3 pF for 67 cells in myopathic group). The average R_S was 1.8 ± 0.1 MΩ for all cells examined. Currents rarely exceeded 1.2 nA, and the maximal voltage error did not exceed 3 mV.

Voltage command pulses were generated by a 12-bit digital-to-analog converter controlled by pCLAMP software (Axon Instruments). APs were elicited by pulses of 2 ms and 90 mV at a rate of 1 Hz. AP measurements were begun 3 minutes after cell membrane rupture, and the steady-state AP duration was measured at 20% (APD₂₀), 50% (APD₅₀) and 90% (APD₉₀) of full repolarization. Depolarization-induced currents included the transient outward K⁺ current (I_TO) and the delayed rectifier outward K⁺ current (I_K), and were elicited by applying steps from a holding potential (V_h) of -40 mV in 10 mV increment up to +60 mV at a frequency of 0.1 Hz.

In recording calcium currents, to minimize the effect of "rundown" 13 , the depolarizing steps were applied from a holding potential which was alternated between -90 mV and -50 mV. Also the steps were applied between 5 and 15 minutes after rupturing the membrane patch. I_Ca,L was measured as inward current during depolarizing 300 ms steps applied from the holding potential of -50 mV in 10 mV increments up to +60 mV. I_Ca,T was separated from I_Ca,L by subtraction of currents elicited by 300 ms depolarizing steps from holding potentials of -50 and of -90 mV in increments of 10-mV up to +60 mV. The calcium currents were measured as the difference between the inward peak and the current remaining at the end of the voltage step.

The background current I_K1 was measured during steps from holding potential of -40 mV in 10 mV increments to test potentials ranging from -20 to -120 mV at a frequency of 0.1 Hz. The holding potential of -40 mV was used to inactivate the sodium channel. For measurement of Na⁺-Ca²⁺ exchange current (NCX), the external solution (in mM) consisted of NaCl 140, CaCl₂ 2, MgCl₂ 1, HEPES 5, and glucose 10 with a pH of 7.4, and contained strophanthidin (4 μM) nitredipine (10 μM) and niflumic acid (15μM). Test steps were applied from a holding potential of -40 mV to test potentials of -100, -80, -60, -40, -20, 0, +20, +40, +60, +80 and +100 mV.

All quantitative data are expressed as mean ± S.E.M. The differences between healthy and myopathic cardiomyocytes were analyzed by one-way ANOVA. The χ² test with Yates' correction of Fisher's exact test was used for the categorical data. A value of p < 0.05 was considered to be statistically significant.

In driven cardiomyocytes, APD₅₀ and APD₉₀ were shorter in myopathic (n = 29) than in healthy hamsters (n = 42). Figure 1A and 1B show typical examples. The difference in APD between the two types of cells was statistically significant (Table 1). Myopathic myocytes had a larger capacitance (Table 1) and thus a larger cell size. Other AP parameters (APA and MDP) were not significantly different between the two groups.

In the absence of electric stimulation, the LA-PV cardiomyocytes could be spontaneously active. As illustrated in Figure 2, spontaneous APs had a slow rate of rise of the upstroke and a conspicuous diastolic depolarization, similar to those of pacemaker cells in the sinoatrial node (SA node). Spontaneously active cardiomyocytes were present in both myopathic and healthy LA-PV areas, but they were more numerous in myopathic hamsters (9/29) than in healthy hamsters (4/42) (χ² = 3.966, p < 0.05).

Triggered rhythms (EAD and DAD) in myopathic LA-PV cardiomyocytes driven at 1 Hz are shown in Figure 3. EAD were present in 5 out of 29 (17%) myopathic and in 4 out of 42 (10%) healthy LA-PV cardiomyocytes. The incidence of DAD was 2/29 (7%) and 8/42 (19%) in myopathic and healthy myocytes, respectively. There were no significant differences in the incidence of EAD and DAD between the two groups. However, it is not readily apparent why EADs and DADs should be present in both normal and myopathic cells.

Selected membrane current traces elicited by depolarizing steps from V_h -90 mV (I) and -50 mV (II, I_Ca,L) in healthy (panel A) and myopathic (panel B) LA-PV myocytes are illustrated in Figure 4. The difference current (I-II, I_Ca,T) arise from the subtraction of current elicited from V_h -90 mV and -50 mV, respectively.

The voltage-dependence of peak I_Ca,L and I_Ca,T density from 12 healthy and 10 myopathic LA-PV myocytes is shown in panel C. The current-voltage relationship of I_Ca,T had a threshold voltage of about -40 mV and a peak at -20 mV. The current-voltage relationship of I_Ca,L had a threshold voltage about -20 mV and a peak at +20 mV. I_Ca,T was activated at more negative voltage than I_Ca,L and its density was three-fold smaller than that of I_Ca,L. The result shows that myopathic myocytes had smaller I_Ca,L than healthy myocytes (-4.6 ± 0.98 versus -7.8 ± 0.75 pA/pF, p < 0.05). The I_CaT was also appeared smaller in myopathic myocytes but the difference between myopathic vs. healthy was not statistically significant.

I_TO was studied with a double-pulse protocol (Figure 5). A 30-ms prepulse from -80 to -40 mV was used to inactivate the sodium channel, followed by a 300-ms test pulse to +60 mV in 10-mV increments. Selected membrane currents in healthy (panel A) and myopathic (panel B) LA-PV myocytes are illustrated in Figure 5. There were no statistically significant differences in voltage-dependent properties of I_TO in myopathic versus healthy hamsters.

Long (1 s) depolarizing steps from V_h -40 to +60 mV in 10 mV increments (see protocol on Figure 6) induced a slowly activating and non-inactivating outward current, consistent with the behavior of delayed rectified K⁺ current in hamster LA-PV myocytes. In addition, there was I_TO with rapid activation kinetics. The voltage-dependence of I_K was comparable in myopathic and healthy hamsters (Figure 6C).

I_K1 was measured in the absence and presence of 1 mmol/L Ba²⁺ as illustrated in Figure 7. The V_h was -40 mV and then I_K1 was measured at test potentials over the range from -20 down to -120 mV (see clamp protocol on top of Figure 7). The difference in the magnitude of I_K1 in control solution and solution containing Ba²⁺ gave the Ba²⁺-sensitive inward rectifier K currents. As shown in Figure 7 panel C, the mean Ba²⁺-sensitive I_K1 was not significantly different between the two groups.

The protocol tested is shown in Figure 8 (top). In Figure 8 V_h was -40 mV to inactivate fast Na channel. For inward NCX currents the cardiomyocytes were hyperpolarized to -100~ -60 mV in 20 mV steps from a V_h of -40 mV. For outward NCX currents, the myocytes were depolarized from V_h -40 to +100 mV in 20 mV increments. Both the inward and outward I_NCX were smaller in the myopathic (panel B) than in the healthy cells (panel A). As shown in the current-voltage relationship of I_NCX (Figure 8, panel C), the difference between the mean outward and inward NCX currents in myopathic cells (n = 5) was significantly smaller than in healthy myocytes (n = 6)

It has been demonstrated in human myocardium that there are myocardial cells in the junction between atrium and PV 17 18 . In isolated guinea-pig PV tissues, infusion of noradrenaline (10^-7 g/ml) can induce spontaneous tachyarrhythmias 19 . Our recent study on PV tissues of dog 1 11 and rabbit 12 13 also demonstrated that PV myocardial sleeves have some cells that can be spontaneously active. In the present experiments, myocytes of myopathic hamsters had automatic activity that was significantly more frequent (31%) than those from healthy hamster (10%,). Thus, the myopathic PVs are more prone to develop automatic arrhythmias.

The increase in automaticity of the myopathic PV cardiomyocytes could be related to one or more of the following ionic mechanisms: an increased I_Ca,T 20 , an enhanced decay in I_K during diastole, a positive shift of the diastolic potential or a negative shift of the threshold potential or a shorter APD 6 16 . The present electrophysiological results suggest that the significant shorter APD (and presumably a shorter refractory period) with an increased slope of diastolic depolarization may be the most likely factors in the abnormal rhythms.

Myopathic PV cardiomyocytes tend to develop EAD and DAD which could initiate repetitive discharge. Yet, the incidences of triggered activities in myopathic myocytes were not significantly higher than those of healthy PV cardiomyocytes. Arrhythmias in myopathic myocytes could also be related to the significant shorter APD₅₀ and APD₉₀ (Table 1), which could facilitate re-entry rhythms through a shorter refractory period. The shorter APs could be related to the smaller calcium currents.

The significantly smaller I_Ca,L could result in a smaller intracellular Ca²⁺ concentration ([Ca²⁺]_i) in myopathic myocytes and therefore be less likely to generate [Ca²⁺]_i overload and transient inward curren I_TI on repolarization from depolarizing steps 6 21 . On the other hand, release of Ca²⁺ from the SR could inactivate the L-type Ca²⁺ channels 22 thus resulting in a smaller I_Ca,L. Further experiments using fluorescent method to measure [Ca²⁺]_i and using whole-cell patch-clamp technique to determine I_TI in myopathic vs. healthy cardiomyoyctes need be carried out to clarify this point.

Hano et al. 23 reported that spontaneous and sporadic ventricular premature contractions (VPC) occurred in 8.3% of Bio14.6 strain cardiomyopathic hamsters, whereas no ventricular arrhythmia was recorded in normal hamsters. Either non-sustained ventricular tachycardia (NSVT) or ventricular fibrillation could be induced in all cardiomyopathic hamsters. In contrast, neither NSVT nor ventricular fibrillation was induced in normal hamsters 23 .

Cold-immobilization of Bio14.6 myopathic hamsters for 2 hr had a lethal effect 24 . There was an obvious heart rate slowing and occasional AF or atrio-ventricular block. In contrast, no ill effects were observed in healthy hamsters subjected to similar stress. Propranolol (but not phentolamine or atropine) prevented the lethal effects of the stress. Higher incidence of automatic activity in the presence of shorter APD and/or refractory period could lead to generation of ventricular arrhythmias 23 but could not explain the atrial arrhythmias 24 . Sakamoto 25 summarized the electrical and ionic abnormalities in the heart of cardiomyopathic hamsters and noted the contradictory results from different or even from the same laboratory. Species difference and ages of animals should be considered in the interpretation of the experimental results. Also, reduced coupling of the myocytes in the PV due to histological changes in extracellular collagen matrix in chronic atrial pacing- induced AF may provide an additional mechanism facilitating repetitive rapid activities 26 .

We have recently reported a study on the mechanisms of arrhythmias in the LA-PV myocardium of mXin-deficient mice induced by brief high frequency electrical drive (30 Hz for 3 sec) 27 in the absence and presence of isoproterenol, strophanthidin and atropine. It was found that automatic rhythms, triggered rhythms and induction of reentrant AF were affected differently by mXinα gene deletion as detected by a MED64 multi-electrode array system 27 . The induction of AF was consistently hindered in mXinα-deficient LA-PV preparations even under conditions that enhance its induction in mXinα+/+ preparations. The mechanisms that prevent the induction of AF in the mXinα-/- preparation appear to involve a decrease in conduction velocity and longer APs: these changes apparently prevent the establishment of reentry paths 27 . However, automatic and triggered rhythms were not suppressed in mXinα-/- preparations.

In contrast to murine cardiomyocytes, the APD of LA-PV cardiomyocytes of myopathic hamster was significantly shorter than that of healthy control (see Figure 2 and Table 1). A shorter AP may favor the generation of reentrant rhythms. Since the incidence of triggered activities in myopathic myocytes was not significantly higher than that in healthy control, other factors such as age-related changes in Na⁺-Ca²⁺ exchanger have been considered 28 29 . However, the present results show that both inward and outward Na⁺-Ca²⁺ exchange current was reduced in the LA-PV cardiomyocytes of myopathic hamsters (Figure 8). Therefore, I_NCX is unlikely to contribute to enhanced automatic rhythms in Bio14.6 hamsters. Further experiments are required to clarify the underlying mechanisms responsible for the arrhythogenesis in myopathic Bio14.6 hamsters.