A precise understanding of how the contraction of the heart – driven by cardiac muscles cells; cardiomyocytes – is paramount to expanding our knowledge of cardiac (patho-)physiology. Most of our knowledge on cardiac dynamics is built upon light microscopic studies of living specimens. Electron microscopy, which exceeds the resolution of light microscopy by far, does not traditionally allow for imaging of living samples, as they have to be preserved, normally accomplished by the use of chemical fixation and dehydration.

Here, we use action potential-synchronised high pressure freezing to allow for the assessment of cardiac contraction with electron-based imaging. We isolated rabbit ventricular cardiomyocytes, high pressure-froze them with milli-second accuracy, and subsequently freeze-substituted, heavy metal-stained, resin-embedded, and cut them into 200–300 nm sections. Then, we employed dual-axis electron tomography using a 300 kV transmission electron microscope, achieving isotropic voxel sizes of (1.2 nm)3 to reconstruct intracellular organelles at specific timepoints during cardiac contraction (Fig. 1). To accelerate data analysis, we developed a machine learning-based workflow that allows us to rapidly extract organelle contours via the surrogate of boundary estimation. This workflow allows for the automated segmentation of sarcoplasmic reticulum, the transverse-axial tubular system, caveolae, microtubule, and mitochondria. Furthermore, we have developed tools to facilitate proofreading and correcting the segmentation. To generate a deeper comprehension of the complex intracellular structure and organisation, we present the results within an interactive virtual reality environment.

High pressure-freezing overcomes usual dehydration and aldehyde-based fixation artefacts, which can distort the image considerably. Our protocol, utilising action potential-synchronised high pressure-freezing combined with three-dimensional, nanoscopic electron tomography and high-throughput automated analysis, will provide key insights into cardiomyocyte ultrastructure and its mechanical modulation.