Phase-targeting rapid-freezing of the beating heart unveils spatiotemporal inhomogeneity of the myocyte sarcomere dynamics

The heart consists of numerous numbers of cardiomyocytes showing coordinated contractions and relaxations as a functional syncytium. However, unknown is to what extent the myocyte sarcomere arrangements differ in each beating phase.

To address this issue, we demonstrated the phase-targeting rapid-freezing of the Langendorff-perfused rat heart to visualize the phase-dependent differences in the distribution of sarcomere length (SL).

Using a home-built cryogen-ejection system, the heart electrically paced at 0.5 Hz was rapidly frozen during peak systole or end diastole by adjusting the timepoint of cryogen exposure to the hearts. Figure 1 shows side-view pictures of the perfused rat heart during rapid-freezing process. The rapidly frozen heart was subsequently fixed with paraformaldehyde (PFA) and acetone under freeze-substitution procedures, then served for immuno-histochemistry of α-actinin to visualize Z-line cycles in sarcomere structures. Fluorescence images obtained by a confocal laser-scanning microscope showed fine arrangements of the Z-line in the myocytes in a subepicardial myocardium (>44,000 µm²/image). The representative sarcomere length of each myocyte was calculated by discrete Fourier transform (DFT) for line-profiles of Z-line cycle.

The phase-targeting rapid-freezing of the beating hearts revealed shorter SLs in the RF-systole heart than those in RF-diastole heart. Figure 2 shows immuno-stained fluorescence images of the hearts rapidly-frozen during systole and diastole. Quantitatively, the SLs of the rapidly frozen heart during peak systole (RF-systole: 1.57 ± 0.12 µm, n = 5 hearts) was significantly shorter than those during end diastole (RF-diastole: 1.92 ± 0.14 µm, n = 5). The SLs of RF-diastole was close to those of the rapidly frozen heart under pharmacological relaxation by 2,3-butanedione monoxime (RF-BDM-diastole: 1.97 ± 0.11 µm, n = 3) and those fixed by PFA-perfusion (PFA: 1.82 ± 0.11 µm, n = 3). The heatmap for the individual SL of the RF-systole heart revealed nearly uniform and shorter SL distributions with some local longer SLs. The heatmap for RF-diastole also showed non-uniform patterns: patchy distributions of short-SL regions within the predominant SL elongation. Such SL spatial inhomogeneity was not seen in RF-BDM-diastole and sole PFA-perfused fixation. Additionally, we demonstrated "snapshots" of spatiotemporally-chaotic SL distributions dynamics in the heart during ventricular fibrillation by rapid-freezing, which was not available by sole PFA-perfused fixation.

Overall, the phase-targeting rapid-freezing of the beating heart provides in-depth behaviors on SLs, which was not obtained by the current spatiotemporally high live imaging so far. Our strategy would contribute to understand precise spatiotemporal changes of cardiac functions.Figure 1For image description, please refer to the figure legend and surrounding text.  Figure 2For image description, please refer to the figure legend and surrounding text.