Comparison of Different Tissue Clearing Methods for Three-Dimensional Reconstruction of Human Brain Cellular Anatomy Using Advanced Imaging Techniques

Abstract

The combination of tissue clearing techniques with advanced optical microscopy facilitates the achievement of three-dimensional (3D) reconstruction of macroscopic specimens at high resolution. Whole mouse organs or even bodies have been analyzed, while the reconstruction of the human nervous system remains a challenge. Although several tissue protocols have been proposed, the high autofluorescence and variable post-mortem conditions of human specimens negatively affect the quality of the images in terms of achievable transparency and staining contrast. Moreover, homogeneous staining of high-density epitopes, such as neuronal nuclear antigen (NeuN), creates an additional challenge. Here, we evaluated different tissue transformation approaches to find the best solution to uniformly clear and label all neurons in the human cerebral cortex using anti-NeuN antibodies in combination with confocal and light-sheet fluorescence microscopy (LSFM). Finally, we performed mesoscopic high-resolution 3D reconstruction of the successfully clarified and stained samples with LSFM.

Keywords: clearing techniques; expansion microscopy; immunofluorescence; light-sheet fluorescence microscopy; optical microscopy.

FIGURE 1

Sample processing. (A) Conceptual diagram of the sample processing pipeline. Human brain slices with a thickness greater than 100 μm were cleared with different clearing protocols (CLARITY, SWITCH, and SHIELD) and then immunolabeled (IF, immunofluorescence). Slices with a thickness less than 100 μm were stained before to perform ExM. All the samples were acquired with different optical techniques. Created with BioRender.com. (B) CLARITY sandwich preparation for hydrogel inclusion. Human brain slices were placed on two coverslips, separated by a 500 μm-thick stainless steel spacer. The sandwich was then filled with a hydrogel CLARITY solution using a syringe and placed in a 50-ml tube dipped with the same solution and incubated at 4°C. After the samples were degassed with nitrogen (N₂) removing the oxygen (O₂). The temperature was increased to 37°C to initiate polymerization. Finally, the embedded sample was extracted from the gel and washed with a clearing solution to remove lipids. Created with BioRender.com. (C–F) Representative images of human brain slices pre-clearing (in PBS) and after refractive index matching in 68% TDE: CLARITY (A), SWITCH (B), and SHIELD (C) and dH₂O (ExM) (F).

FIGURE 2

NeuN antibodies validation using confocal microscopy. (A–C) Representative confocal images of human brain slices stained using anti-NeuN antibody (ABN91; left): SWITCH (A), SHIELD (B), and CLARITY (C) samples, respectively. White arrows highlight three different stained neuronal bodies. On the right, autofluorescence signal of lipofuscin (white arrows). Excitation light, 568 nm; laser power, 1 mW. Objective lens, 60×; NA, 1.4. Scale bar = 50 μm. (D,E) Representative confocal images of SWITCH (D) and SHIELD (E) clarified human brain slices stained using ABN91, ARG 10712, and 26975-1-AP anti-NeuN antibodies. Scale bar = 50 μm.

FIGURE 3

Representative LSFM images of clarified human brain slices labeled with different NeuN antibodies. Mesoscale reconstruction of SWITCH- (A) and SHIELD- (B) processed slices stained with anti-NeuN antibody ABN91 with Alexa Fluor 568. Excitation light 561 nm, laser power 5 mW. Maximum intensity projection (MIP) within 60 μm at 150 μm depth. Scale bar = 1 mm. (C,D) Lateral reslicing of SWITCH (C) and SHIELD (D) processed samples with corresponding normalized plot intensity profiles (black line: mean; light-red outline: standard deviation). Scale bar = 250 μm. (E) The graph shows the neuron quantification in the prefrontal cortex of two subsequent slabs of the same subject using SWITCH and SHIELD clearing techniques. The counting was performed using 3 ROIs selected at different penetration depths (0–160, 170–330, and 340–500 μm) for 6 random stacks (total 18 ROIs) in the gray matter. Data are reported as mean ± SD and two-sample t-test performed (**P < 0.01). (F) The column plot shows the signal to background ratio of neurons at different depths (0–50, 225–275, and 450–500 μm). It was not possible to identify labeled neurons in the center of the SHIELD-slabs, ND, not detectable.

FIGURE 4

Representative LSM Broca’s area images. (A) Broca’s area images of human brain slices in PBS (pre-clearing) and after refractive index matching in the 68% TDE/PBS solution. (B) SWITCH-processed slice stained with anti-NeuN antibody. MIP of 30 μm at 250 μm depth. Scale bar = 1 cm. (C) High-resolution insert corresponding to the white square (3 mm × 3 mm) in (B). Scale bar = 1 mm.

FIGURE 5

NeuN immunostaining validation using ExM. (A) Confocal images of pre- and post-expanded human cortex immunostained for NeuN (antibody ABN91) with Alexa Fluor 488. Objective lens, 60×; NA, 1.4; excitation light, 488 nm. Signal-to-noise ratio comparison between the tissue transformation protocol SWITCH, pre-expansion (classic immunofluorescence protocol), and post-expansion specimens. A significant improvement is observable in the post-expanded samples, imputable to the dilution of the non-specific signal and reduction of the autofluorescence. The choice of the fluorescent dyes and the acquisition setting were optimized for each treatment and clearing procedures. Pre-expansion scale bar = 10 μm; post-expansion scale bar (40/4) = 10 μm. (B) Expansion factor characterization using the sample thickness before (∼100 μm) and after expansion (∼400 μm) acquired by LSFM (N = 10). Pre-expansion scale bar = 100 μm; post-expansion scale bar (400/4) = 100 μm. (C) Maximum intensity projection of post-expanded human superior frontal cortex using our custom-made LSFM. Excitation light, 488 nm; power, 5 mW. Scale bar = 1 mm. The white arrows show the sample thickness used for quantifying the expansion factor.