During cardiomyocytes membrane displacement measurements, the pulsated cardiomyocyte pushed up the cantilever on it when systole. However, the cardiomyocyte should be considered as a soft material. We have to optimize the force applied on the cardiomyocyte in case the force is so high that the cardiomyocyte pulsation suppressed by our cantilever. Therefore, we measured the membrane displacement at different forces applied on the cardiomyocyte. In Figure 9, by changing the force applied on the cardiomyocyte, we can see the membrane displacement was suppressed by higher loading on it, the membrane displacement dropped down when the loading force was 20 nN. The feed-back loop in SPM keeps the separation between the cantilever and the cell membrane, but still a limit of SPM feed-back loop there. By increasing the i gain and p gain in the controlling program, we can trace the membrane movement more accurately. Under the maximum integral gain and proportional gain without cantilever oscillation, there are still small cantilever fluctuations. In order to measure isotonic contraction of cardiomyocytes, the force applied on the cell should be higher than the fluctuation. The minimum force we applied on the cardiomyocyte to ignore the fluctuation was about 4 nN. The measured displacement before 4 nN due to instrumental error that the cantilever couldn’t trace the membrane very well just like false engaging, so we saw the membrane displacement was lower than the displacement at 4 nN.
30
0 10 20 30 40 50
0.9 1.2 1.5
Displacement ( μ m)
Setpoint (nN)
Figure 9 Dependence of membrane displacement on the loading force
When the loading force is higher than 20 nN, the measured cardiomyocyte membrane displacement is smaller due to overloading force applied on the cardiomyocyte.
However, when the loading force is near 4 nN, the measured cardiomyocyte membrane displacement reaches a maximum value.
4-3 Effect of density to pulsation behaviors
For monitoring the membrane movement of cardiomyocyte, the constant force applied on it was characterized. Reasonable explanation for the suppressed membrane displacement can be related to the elastic component in cardiomyocytes. When the force increased to 20 nN, we saw the pulsation displacement started to decline, this result also consistent with previous work (the loading force up to 10 nN had no detectable influence on the membrane displacement).30
We first examined the time traces of pulsation for single cells cultured in different cell densities. For a confluent layer of cells, the recorded time trace is very regular indicating that cells cultured in a confluent layer (2 × 104 cells/dish, Figure 10(A)) exhibit synchronous and stable pulsation (Figure 11(A)). In contrast, the pulsation behavior for cells in a sub-confluent layer (104 cells/dish, Figure 10(B)) is irregular (Figure 11(B)); the cells stop pulsing and remain in a quiescent state occasionally exhibiting a characteristic behavior resembling that of arrhythmia. Interestingly, the time trace shows that the pulsation amplitude does not seem to vary significantly despite that the pulsation frequency seems to vary drastically from time to time. This result demonstrates the capability of SPM to delicately characterize the pulsation behavior of cardiomyocytes cultured in different cell densities. To gain insight of the dynamic aspect of these pulsation behaviors, we employ short-time Fourier transform (STFT) to obtain the time-varying frequency of cardiomyocyte pulsations. As expected, the spectrogram of pulsations for cardiomyocytes embedded in a confluent layer shows a characteristic frequency with no appreciable change along with time (Figure 12(A)).
In contrast, the spectrogram of pulsations obtained from cells embedded in a sub-confluent layer carries no distinct characteristic frequency (Figure 12(B)). As demonstrated, the spectrogram obtained from STFT analysis clearly illustrates the
32
dynamic behavior of pulsation. We will show that this intuitive while informative representation of a dynamic change is particularly useful to assess the function of cardiomyocytes.
We control the cell density of sub-confluent layer and confluent layer of cardiomyocytes. The tight junction between cells only achieved in confluent layer of cardiomyocytes, synchronize contractions were observed in bright field image.
However, the sub-confluent layer of cardiomyocytes contracted in similar amplitude with unstable pulsation frequency compared with confluent layer. We can see the contractility of each cardiomyocyte in sub-confluent layer is almost the same because of the same physiological condition, only the irregular frequency caused by bad junctions. For the drug testing model, the pulsation behavior of confluent layer of cardiomyocytes is more suitable.
20 μm 20 μm
(A) (B)
Figure 10 Bright field images of 6-day cultured chicken embryo cardiomyocytes under two plating density
(A) The plating density is 2 × 104 cells/dish; tight junctions were formed between cardiomyocytes. (B) The plating density is 104 cells/dish; junctions between cardiomyocytes are rare.
34
102 104 106 108 110 112 0.0
Figure 11 The effect of plating density control on the pulsation behavior represented by time trace
(A) We can see the pulsation displacement and pulsation frequency of confluent layer of cardiomyocytes is 0.562 ± 0.003 μm and 1.658 ± 0.010 Hz, respectively.
(B) Time trace of sub-confluent layer pulsation behaviors, with small differences in
displacements between each pulsation in 600 seconds. We can see the pulsation displacement and pulsation frequency is 0.570 ± 0.014 μm and 1.561 ± 0.065 Hz, respectively.
36
(B) (A)
Figure 12 The effect of density control to cardiomyocytes pulsation behaviors represented by spectrogram.
(A) Time trace of confluent layer pulsation behaviors showed very stable pulsation behaviors in displacements and frequencies in 600 seconds. The spectrogram of cardiomyocytes embedded in a confluent layer shows a characteristic frequency with
no appreciable change along with time. (B) The spectrogram of cardiomyocytes embedded in a sub-confluent layer carries no distinct characteristic frequency in 600 seconds, which showed the time-varying pulsation frequencies, which pulsation frequencies were irregular.
38