Cardiac Light-Sheet Fluorescent Microscopy for Multi-Scale and Rapid Imaging of Architecture and Function

Light Sheet Fluorescence Microscopy (LSFM) enables multi-dimensional and multi-scale imaging via illuminating specimens with a separate thin sheet of laser. It allows rapid plane illumination for reduced photo-damage and superior axial resolution and contrast. We hereby demonstrate cardiac LSFM (c-LSFM) imaging to assess the functional architecture of zebrafish embryos with a retrospective cardiac synchronization algorithm for four-dimensional reconstruction (3-D space + time). By combining our approach with tissue clearing techniques, we reveal the entire cardiac structures and hypertrabeculation of adult zebrafish hearts in response to doxorubicin treatment. By integrating the resolution enhancement technique with c-LSFM to increase the resolving power under a large field-of-view, we demonstrate the use of low power objective to resolve the entire architecture of large-scale neonatal mouse hearts, revealing the helical orientation of individual myocardial fibers. Therefore, our c-LSFM imaging approach provides multi-scale visualization of architecture and function to drive cardiovascular research with translational implication in congenital heart diseases.


Fig. S1
The schematic diagram of c-LSFM modality Fig. S2 4-D zebrafish beating heart reconstruction methods

Fig. S3
Post-image processing of adult zebrafish heart with resolution enhancement by different deconvolution techniques.

Fig. S4
The P1 neonate mouse hearts before and after BABB clearing Fig. S5 Synchronization comparison before and after.

Fig. S6
The imaging comparison between c-LSFM and confocal microscope using 120 dpf zebrafish hearts.

Fig. S7
Comparison of 4-D synchronized images with a combination of 3 different parameters Table S1 Analysis of cardiac mechanics in 4dpf zebrafish from 4-D in vivo imaging Video S1 The high speed, high contrast imaging of live embryonic zebrafish heart Video S2 Cardiac cycle before temporal interpolation Video S3 Cardiac cycle after temporal interpolation Video S4 The 4D reconstruction of live embryonic zebrafish heart Video S5 The image acquisition process of c-LSFM Video S6 The high resolution reconstruction of digital mouse heart using Amira Furthermore, eliminating the sealing of the objectives into a water chamber additionally benefits the frequent changes in immersing mediums for different samples. Therefore, the c-LSFM modality we specially designed is very efficient in performing trans-scale cardiac light sheet imaging conveniently. After back-projection, we compared samples at a same spatial location but from different periods and evaluated each of the period hypotheses. The best hypotheses were selected accordingly. Relative shift determination aimed at aligning the starting sample of each individual image sequence. The heart may not be in the same contraction state at the beginning of all sequences when we start taking images at each z layer. Starting from a number of relative shift hypotheses, we adopted a quadratic cost function to measure the alignment. The cost function is calculated via measuring the similarity between two hypothetically aligned images from two adjacent image sequences with respect to the relative shift hypotheses. By maximizing the alignment, we select the best possible relative shift hypothesis for each image sequence with respect to the other sequences.
Absolute shift determination targeted to obtain the absolute shift of each individual image sequence with respect to the first sequence. In the previous step, relative shift between any close-by image sequences are obtained. We recursively calculate relative shift between the current image sequence and an early sequence until obtaining the relative shift with respect to the first image sequence. The above process is applied to every image sequence, and all such processes can be compactly implemented by one matrix multiplication. However, it also generates image discontinuity, likely due to the application of wiener filter. The CGLS algorithm appears to be mild, generating the least degree of deblurring. Of the three resolution enhancement algorithms, the MRNSD method provides optimal trade-off between the resolution enhancement and information preservation. We optically cleared the day 1 neonatal mouse heart with 2 hours serial ethanol dehydration followed by 2 hours benzyl alcohol-benzyl benzoate clearing.
Compared to the raw hearts that were completely opaque before clearing (a), the treated hearts showed significantly reduced scattering and became highly translucent on a scale board.   identical image quality to that of 100fps. Therefore, we selected (d) as the optimal combination for 4-D synchronized imaging parameters. Scale bar = 10m. Table S1. Analysis of cardiac mechanics in 4dpf zebrafish from 4-D in vivo imaging.
Video S1 shows the high speed, high contrast imaging of a live embryonic zebrafish heart. The fast interaction of blood flow (ds-red) and beating cardiac muscle (GFP) was captured and synchronized.
Video S2 shows the beating heart of a live embryonic zebrafish before temporal interpolation.
Video S3 shows the beating heart of a live embryonic zebrafish after temporal interpolation.
Video S4 shows the 4-D reconstruction of a live embryonic zebrafish heart. Video S5 shows the acquisition process of day 1 neonatal mouse heart using c-LSFM.
The frame rate was as fast as 20 fps under full resolution, which enabled 3D scanning of an entire heart in around 30 seconds.
Video S6 shows the high resolution, digital neonatal mouse heart reconstructed using Amira. We can easily access an area of interest with a cellular level resolution.