Efficient 3D imaging and pathological analysis of the human lymphoma tumor microenvironment using light-sheet immunofluorescence microscopy

Abstract

Rationale: The composition and spatial structure of the lymphoma tumor microenvironment (TME) provide key pathological insights for tumor survival and growth, invasion and metastasis, and resistance to immunotherapy. However, the 3D lymphoma TME has not been well studied owing to the limitations of current imaging techniques. In this work, we take full advantage of a series of new techniques to enable the first 3D TME study in intact lymphoma tissue. Methods: Diverse cell subtypes in lymphoma tissues were tagged using a multiplex immunofluorescence labeling technique. To optically clarify the entire tissue, immunolabeling-enabled three-dimensional imaging of solvent-cleared organs (iDISCO+), clear, unobstructed brain imaging cocktails and computational analysis (CUBIC) and stabilization to harsh conditions via intramolecular epoxide linkages to prevent degradation (SHIELD) were comprehensively compared with the ultimate dimensional imaging of solvent-cleared organs (uDISCO) approach selected for clearing lymphoma tissues. A Bessel-beam light-sheet fluorescence microscope (B-LSFM) was developed to three-dimensionally image the clarified tissues at high speed and high resolution. A customized MATLAB program was used to quantify the number and colocalization of the cell subtypes based on the acquired multichannel 3D images. By combining these cutting-edge methods, we successfully carried out high-efficiency 3D visualization and high-content cellular analyses of the lymphoma TME. Results: Several antibodies, including CD3, CD8, CD20, CD68, CD163, CD14, CD15, FOXP3 and Ki67, were screened for labeling the TME in lymphoma tumors. The 3D imaging results of the TME from three types of lymphoma, reactive lymphocytic hyperplasia (RLN), diffuse large B-cell lymphoma (DLBCL), and angioimmunoblastic T-cell lymphoma (AITL), were quantitatively analyzed, and their cell number, localization, and spatial correlation were comprehensively revealed. Conclusion: We present an advanced imaging-based method for efficient 3D visualization and high-content cellular analysis of the lymphoma TME, rendering it a valuable tool for tumor pathological diagnosis and other clinical research.

Keywords: 3D spatial analysis.; light-sheet microscopy; lymphoma tumor microenvironment; three-dimensional imaging; tissue clearing.

Figure 1

Overview of lymphoma TME 3D imaging and analysis pipeline, which contains four main steps. Step 1: collection of fresh lymphoma tissue samples, which were fixed immediately with 4% PFA and stored at 4℃. Step 2: immune-labeling of samples, wherein samples were first pretreated with methanol series concentration for dehydration, 5% H2O2 for bleaching, and methanol series concentration for rehydration, followed by antibody labeling. Step 3: tissue optical clearing, which included tert-butanol dehydration, DCM delipidation, and matching for BABB-D4 until they became transparent. Step 4: 3D imaging of labelled transparent tissues with a light sheet microscope and image-based pathological analysis of the TME.

Figure 2

Comparative performances of four tissue clearing methods for human lymph node specimens. (A) Protocol and clearing time of uDISCO, iDISCO+, SHIELD, and CUBIC. (B) Photographs of 1-mm thick human lymph node tissue blocks before and after clearing using uDISCO, iDISCO+, SHIELD and CUBIC. The scale of the grid is 1 mm × 1 mm. (C) Comparison of achievable imaging depth of four clearing methods. Scale bar = 200 μm.

Figure 3

Antibody panel for targeting diverse lymphoma immune cells and LSFM images of each type of labeled immune cells. (A) Screened antibody panel for tagging the interested cell subtypes in the lymphoma TME. (B) The x-y plane images of the cleared lymphoma samples showing the spatial distributions of available 8 antibodies. Scale bars are 200 μm in the left wide-view images, and 20 μm in the right magnified vignettes, respectively.

Figure 4

3D visualization of the immune TME in lymphoma at single-cell resolution. (A) The schematic of our home-built light-sheet fluorescence microscope suited for high-speed imaging of cleared thick samples. (B) Control diagram of the camera and motorized stages of the microscope during imaging (C) Volume renderings of the human lymphoma immune TME. A lymphoma specimen was divided into four parts with each part labelled with different cell subtypes. ① Tregs and T cells: CD8, red; FOXP3, green. Scale bar = 300 μm. ②Monocytes, neutrophils and T cells: CD14, blue; CD15, red; CD3, green. Scale bar = 500 μm. ③ Macrophages, TAMs and T cells: CD68, blue; CD163, red; CD3, green. Scale bar = 400 μm. ④ B cells, T cells and Ki67: CD20, blue; CD8, red; Ki67, green. Scale bar = 300 μm. (D) Close-up xy and yz planes of diverse antibodies-labeled cell subtypes (CD3, CD8, CD20, CD68, CD163, CD14, CD15, FOXP3 and Ki67). Scale bars = 50 μm in all images.

Figure 5

LSFM image-based data processing flow for analyzing the spatial correlations in the lymphoma TMEs. (A) Post analysis pipeline of the multiplexed LSFM images. (B) The density (cells per mm³) of immune cell types in lymph nodes from three donors with RLN, DLBCL and AITL. i) B cells and T cells: CD20, blue; CD8, red. ii) Tregs and T cells: FOXP3, green; CD8, red. iii) Macrophages, TAMs and T cells: CD68, blue; CD163, red; CD3, green. iv) Monocytes, neutrophils and T cells: CD14, blue; CD15, red; CD3, green. (C) The number of T cells around each immune cell was plotted as a function of the radial distance from the immune cell. Different color bars represent different diseases with blue for RLN, green for DLBCL, and red for AITL.

Figure 6

Spatial analysis of the colocalizations among diverse immune cells in the lymphoma TME. (A) Microenvironment images directly showing the colocalization among the specific immune cells in the RLN, DLBCL and AITL. The yellow circles indicate the cell colocalizations including CD8 (red) and CD20 (blue); CD8 (red) and FOXP3 (green); CD3 (green) and CD68 (blue); CD3 (green) and CD163 (red); CD3 (green) and CD14 (blue); and CD3 (green) and CD15 (red). (B) The boxplots revealing the differences of immune cell colocalizations in RLN, DLBCL, and AITL. (C) The comparative proliferative capacities of B cells and T cells in RLN, DLBCL, and AITL.