Multiplex, quantitative cellular analysis in large tissue volumes with clearing-enhanced 3D microscopy (Ce3D)

Abstract

Organ homeostasis, cellular differentiation, signal relay, and in situ function all depend on the spatial organization of cells in complex tissues. For this reason, comprehensive, high-resolution mapping of cell positioning, phenotypic identity, and functional state in the context of macroscale tissue structure is critical to a deeper understanding of diverse biological processes. Here we report an easy to use method, clearing-enhanced 3D (C_e3D), which generates excellent tissue transparency for most organs, preserves cellular morphology and protein fluorescence, and is robustly compatible with antibody-based immunolabeling. This enhanced signal quality and capacity for extensive probe multiplexing permits quantitative analysis of distinct, highly intermixed cell populations in intact C_e3D-treated tissues via 3D histo-cytometry. We use this technology to demonstrate large-volume, high-resolution microscopy of diverse cell types in lymphoid and nonlymphoid organs, as well as to perform quantitative analysis of the composition and tissue distribution of multiple cell populations in lymphoid tissues. Combined with histo-cytometry, C_e3D provides a comprehensive strategy for volumetric quantitative imaging and analysis that bridges the gap between conventional section imaging and disassociation-based techniques.

Keywords: histo-cytometry; immune system; quantitative microscopy; tissue clearing.

Fig. 1.

C_e3D attains tissue transparency while retaining reporter protein fluorescence and capacity for immunolabeling. (A) Images of fixed LNs were acquired before (PBS) and after C_e3D treatment. (B) LNs were stained with the indicated antibodies, cleared with C_e3D, and imaged by confocal microscopy, with the zoom-in panel (Bottom) demonstrating the capacity to resolve individual cells. (C) Images of various murine organs were acquired before and after C_e3D treatment. (D) Average cellular fluorescence for CD11c-YFP–expressing cells across indicated imaging depths was quantified for PBS (black) or C_e3D (red) -treated tissues. Laser power ramping and signal attenuation correction was not used for quantitative comparison of fluorescence detection. Images represent at least two independent experiments.

Fig. 2.

Large volumetric C_e3D microscopy of diverse organs. (A) Segments of the proximal small intestine were stained with the indicated antibodies, cleared and imaged, allowing visualization of the spatial relationships between the mucosal epithelium (EpCAM⁺, cyan), T cells (CD3⁺, red), and lymphatics (Lyve-1⁺, yellow). (B) C_e3D-processed 1-mm Confetti reporter brain slices were imaged, allowing visualization of neuronal processes and blood vessels. (C) GFAP-stained CD11c-YFP brain slices demonstrate compatibility of reporter and immunolabeling-based fluorescence, as well as preservation of cell morphology with C_e3D. (D) Cleared DsRed reporter brain tissues allow visualization of complex vascular networks. Cytokeratin-stained lungs from CD11c-YFP mice allow 3D visualization of bronchial epithelial cells and dendritic cells via light-sheet (E) or confocal (F) microscopy. (G) Visualization of periosteal blood vessels within actin-DsRed C_e3D-cleared femurs. Images represent at least two independent experiments.

Fig. 3.

Volumetric segmentation pipeline for quantitative tissue analysis. (A) Cell membrane stains were normalized by intensity and summed together to yield the sum channel. This channel was then inverted to create an inverse sum channel and used to create 2D skeletons in the XY, XZ, and YZ planes. These skeletons were then subtracted from the inverse sum channel to enhance the separation between cells. This enhanced image was then segmented to create individual cell objects. (B) The statistics for the cell objects were exported into FlowJo for cellular phenotyping and population gating. (C) Positional information on gated cell populations was used to quantitatively visualize cellular position. Data are representative of at least three independent experiments.

Fig. 4.

C_e3D permits volumetric multiplex imaging and quantitative histo-cytometry. (A) LNs were stained with the indicated antibodies, cleared, and imaged with a 20× 0.7 5 N.A. objective. (B) Cell membrane signals were used to create cell objects, which were then exported into FlowJo. The 3D organ reconstruction in FlowJo is presented and compared with the original image. (C) Cell object statistics were used for population phenotypic analysis and gating of blood endothelial cells (BECs), lymphatic endothelial cells (LECs), macrophages (Macs), as well as B and T cells. (D) Gated cell populations from C were displayed on an XY positional plot for the Z-dimension gates defined in B (red rectangles) and compared with the same Z planes from the original image. Images are representative of at least two independent experiments.

Fig. 5.

C_e3D histo-cytometry of closely adjoining cells in large tissue volumes. (A) LNs were stained with the indicated antibodies, cleared, and imaged with a 40× 1.3 N.A. objective (Top). A zoom-in image, corresponding to a single virtual Z section within the area highlighted by the white box (Top), demonstrates the visual clarity of various lymphocyte subpopulations (Bottom). (B) All imaged cells were segmented and their statistical information was exported into FlowJo for quantitative tissue visualization, as well as (C) population phenotypic analysis and gating. (D) Indicated T-cell populations were displayed as pseudocolor density XY positional plots, with the B cells superimposed (gray density gradients) to highlight regionalized cell distribution. Zoom-in image demonstrates an area enriched in CD4⁺CD25⁺ cell clusters, with the original microscopy image presented below. Images are representative of at least two independent experiments.