肿瘤-免疫微环境的高重数3D成像

Solid tumors evolve in competition with the immune system and must attract resources such as nutrients and oxygen to proliferate and invade normal tissue. This results in complex spatial distribution of tumor, immune, and stromal cell types within the tumor microenvironment (TME). Regulatory interactions among these cells involved in tumor progression and response to therapy are found on a wide range of length scales from subcellular membrane domains (involved in autocrine signaling between PD1 and PDL1 for example) to extended structures spanning thousands of cell diameters such as lymphoid aggregates and tumor invasive boundaries. Highly multiplexed tissue imaging (spatial proteomics), in combination with other single cell methods, promises to reveal the composition and regulation of these TME features in molecular detail but the great majority of data collected to date, including all classical histopathology based on hematoxylin and eosin (H&E) and immunohistochemistry (IHC), involves thin 2D tissue slices. We have recently demonstrated that no cells in such 5 micron tissue section are intact and that this substantially impacts cell type identification and interaction analysis (Yapp...Sorger et al (2025) NatMeth, PMID: 41023436). We will now describe the development of a suite of computational and experimental methods for performing 3D image-based profiling of human tumor sections at different resolutions and spatial scales ranging from ~50 micron thick (with confocal imaging) to 1 mm thick (with cyclic light sheet microscopy). Thick section 3D cyclic immunofluorescence (CyCIF) makes it possible to map cell-cell interactions based on the proximity of cell membranes rather than simply the location of nuclei, as commonly done in spatial tissue analysis. Moreover, by performing 50-plus marker, high resolution (below 200 nm), thick section CyCIF we are able to infer dynamic biological processes using fixed time data and trajectory analysis. Chief among these processes is T-cell mediated cytotoxicity as assayed by formation of immune synapses and polarized secretion of granzyme and/or granulysin containing vesicles toward tumor cells. To date, light sheet fluorescence microscopy (LSFM) has been applied primarily to the analysis of tissues in model organisms (particularly mice and fish) with structures of interest labelled either by fusion to fluorescence proteins or by injecting dyes. We have now developed methods to perform cyclic LSFM on human FFPE tissue, enabling large multi-cellular assemblies to be studied in detail. We will describe an analysis of secondary and tertiary lymphoid structures and their roles in programming B and T cells for anti-tumor immunity. Finally, we will describe how different types of 3D profiling can be combined to span the diversity of spatial scales and regulatory interactions within the human TME.