Fig. 1. Flow chart of our proposed pipeline. (a) Preparation of a macaque brain sample. After viral injection, the macaque brain was immersed in the PNAGA solution. Then a hydrogel was formed across brain tissues after fully permeation and polymerization. (b) The imaging system used in our research. The hydrogel embedded brain was fixed in the sample holder on a 3D-stage. A homemade vibratome was used to cut the sample surface. Two-channel line-scan confocal microscopy equipped with sCMOS was used for high-speed data acquisition. (c) Macaque brain image processing. After illumination correction, the acquired stripes of one imaging plane were stitched to form one image and stored in TIFF format. Then, images were transformed to TDat format for big data processing and further image visualization.

Fig. 2. Preservation of brain tissue macroscale and microscale structures with PNAGA-hydrogel embedding. (a) Reaction process of NAGA crosslinking with tissue, NAGA monomer and crosslinker MBA polymerized to form a 3D hydrogel network. (b) Size changes of the brain tissue embedded with PNAGA-hydrogel and PAM-hydrogel after immersed in water. PAM, polyacrylamide (c) Brain size changes during PNAGA-hydrogel embedding measured with micro-CT (n = 6 each) at different time points. Before embedding (0 h), during permeation (6 h), after polymerization (24 h), and after being immersed in water for 1 d, 2 d, and 14 d, respectively. (d) Morphology of neurons labeled with GFP protein and vessels labeled with DyLight 594 before and after PNAGA-embedding. Left, middle and right rows correspond to samples acquired with a confocal microscope before, after PNAGA embedding, and overlaid, respectively. (e) Mean distance changes between neurons-neurons, vessels-vessels, and neurons-vessels measured before and after PNAGA-hydrogel embedding (n = 12 per each tissue structure from  3 different brain slices). (f) Quantitative statistics of cell body volume, branch distance to the cell body, and branch angle changes (n = 12 per each tissue structure from  3 different brain slices). (g) Confocal images of PNAGA-hydrogel embedded mouse brain tissues stained with PI and DAPI at different depths and various staining times. The sub-images in the first line indicate the sample surface. Each sub-image in a raw file is 1 lm in depth away from the nearby two sub-images. (h) Uniform PI staining of the sagittal plane of intact macaque brain hemisphere at different imaging timepoints (upper) and enlarged images of the corresponding cortex (below). D1, D9, D18, D27, and D36 represent images acquired at the 1st, 9th, 18th, 27th, and 36th day. Scale bar, (b), 2 mm; (d, g), 50 lm; (h), 5 mm/1 mm. (b–g) correspond to mouse data, (h) to macaque.