The Mouse ECG and Heart Rate Variability: Methodological Considerations

Abstract

Decreased heart rate variability (HRV) is predictive of various forms of heart disease and may therefore be an important target for genetic and pharmacological studies in mouse models. For decades, ECG recordings have been made in anesthetized mice, despite the varied depressant affects of the anesthetic agents on autonomic regulation of heart rhythm. Since 1993, telemetric devices have been surgically implanted in anesthetized mice for long-term recording of ECGs and HRV in post-operatively freely moving mice.1 We recently developed a non-invasive technique for rapidly screening ECGs and HRV in conscious mice, which does not require anesthetic or surgery. The several techniques being applied in phenotyping mouse models of human diseases are resulting in disparity in the reporting of ECG indices, including PR, QRS, and QT interval durations. Although the anesthetic agent or surgery trauma can largely influence these values, the sampling frequency of the recording system can also affect the interpretation of measurements. To demonstrate this latter point, we compared results of analyses of ECGs acquired non-invasively in conscious mice at either 1,000 samples per second (1 kHz) or 2,000 samples per seconds (2 kHz).

Methods   Figure 1.ECG signal acquired in a mouse with sample rate of 1 kHz (left) or 2 kHz (right).   Three C57BL/6 mice were obtained from The Jackson Laboratory. The AnonyMOUSETM ECG screening system was used to record ECGs.2 No surgery was required. Data were acquired at either 1 kHz or 2kHz for at least 2 seconds to provide equivalent continuous recordings of 20 to 30 beats. e-MOUSETM was used to interpret the signals.2 Time domain indices of HRV were calculated as described previously.3

Results

Discussion

It is now generally accepted that the murine heart rate ranges from 400 to 800 bpm1, that the levels are dependent on the activity of the mouse and time of day1,4, and that there are significant strain, age, and gender differences2. HRV appears to be as important in mice as it is in human, reflecting the dynamic interplay between perturbations to cardiovascular function and the response of the cardiovascular regulatory systems to these perturbations.5 These beat-to-beat fluctuations reflect the influences of both the sympathetic and parasympathetic limbs of the autonomic nervous system. Transgenic mice with cardiac GSalpha overexpression were found to have depressed HRV, due to enhanced sympathetic regulation.4 HRV was diminished in mice deficient in the muscarinic-gated potassium channel IKACh, indicating that IKACh plays a significant role in parasympathetic control of heart rate.6 Here, we demonstrate that the sampling frequency of data collection can significantly affect determination of HRV, and possibly accurate calculation of other ECG indices. Since it is generally accepted that the murine heart depolarization/repolarization process is approximately 10 times more rapid than humans, then it is reasonable to assume that the QRS interval duration would be approximately 1/10th that observed in humans, or approximately 8 to 12 ms. Therefore, a sampling frequency of 1kHz may provide as few as 8 data points to describe ventricular depolarization, which may be insufficient when examining the effects of genetic mutations or new pharmacological agents. Figure 1 illustrates the better resolution of the ECG signal at the higher sampling rate. Table 1 illustrates how the lower sampling rate may overestimate time-domain indices of HRV by less perfectly identifying the peaks of the QRS complexes of the individual signals, and thereby increase the standard deviation of measured interval durations. Examination of ECGs acquired at 2 kHz non-invasively in conscious dystrophin-deficient mice allowed us to recently identify a novel cardiac phenotype in a mouse model of Duchenne muscular dystrophy,7 which may have gone unnoticed had anesthetics or surgical trauma been employed.

References

1. Kramer, K. et al. 1993. J. Pharmac. Toxic. Methods 30:209-215.
2. Chu, V. et al. 2001. BMC Physiology 1:6.
3. Germann, J. et al. 2000. Am. J. Physiol. 279:H733-H740.
4. Uechi, M. et al. 1998. Circ. Res. 82:416-423.
5. Appel, M.L. et al. 1989. J. Am. Coll. Cardiol. 14:1139-1148.
6. Kovoor, P. et al. 2001. J. Am. Coll. Cardiol. 37:2136-2143.
7. Chu, V. et al. 2002. Muscle and Nerve, in press.