Multiscale Analysis of Immune Responses
National Institute Of Allergy And Infectious Diseases
Investigators
Linked publications, trials & patents
Abstract
To extend our Histo-cytometry we have moved in two directions. First, we have developed a method (IBEX) that permits performing rapid iterative 4-10 color fluorescent immunohistochemistry on sections to attain images with >60 markers and have developed novel methods (RAPID and SPACE) to computationally analyze the complex data emerging from this method. To examine tissue volumes rather than sections, we have developed a novel tissue clearing method called Ce3D and extended this to multiplex imaging in Ce3D-IBEX, achieving up to 25 parameters in tissues as thick as 300u. We have developed methods that permit detection of the oligo-conjugated antibodies used in the CITE-seq technique, facilitating direct correlation of the imaging data with RNA seq studies, while also providing an alternative to our iterative bleaching method that is especially suited to volume imaging. Finally, we have incorporated a sensitive enzymatic amplification step into these methods to permit imaging of severely fixed material from BSL4-level infected sources, while adapting all these tools to more rapid imaging using new instruments that incorporate computational clearing. These LBS-developed imaging technologies methods (Histo-cytometry, IBEX, Ce3D, Ce3D-IBEX) are now being employed in multiple distinct mouse tumor models (breast, pancreatic, lung) to explore the detailed spatial organization of the tumor micro-environment and the changes that occur with immunotherapeutic intervention. By examining multiple different tumors in different tissues, we are beginning to develop insights into what aspects of immune cell presence / spatial organization are unique to a particular malignancy and which represent common features across tumor types. Our methods provide a much more comprehensive analysis of the organization of tumor cells, stromal elements, and immune cells than conventional pathology of immunohistochemical methods, and are especially valuable given the disorganized mature of tumors such that single tissue sections or limited parameter analyses fail to reveal larger scale patterns or variations in different regions of the tumor that may be criterial for understanding the differential response among patients to immunotherapeutic interventions. Preliminary data show clear differences in the localization of distinct T cell subsets within various tumors, with CD8 T cells and CD4 T cells (both conventional cells and Tregs) often located quite differently from the CD8 T cells. These patterns also change with administration of checkpoint immunotherapy, and the phenotypic state of the cells is also dramatically altered when various forms of immunotherapy are employed. Similar asymmetric localization of myeloid cell subsets, such as dendritic cells, monocytes, TAM, and so on, is also evident and also changes dramatically when immunotherapy is used. We are now using sophisticated spatial analytic methods, some in hand and others still under development, to better understand these patterns and how they relate to whether immunotherapeutic treatment is effective or not. These studies are not just of material from mouse experimental models, but in the context of the NIAID-NCI Center for Advanced Tissue Imaging (CAT-I), include analysis of samples from humans with various malignancies, including but to limited to follicular lymphoma (FL), mesothelioma, and lung adenocarcinoma. The work on FL has revealed intriguing relationships between acellular matrix and malignant B cells as well as between mutationally added oligosaccharides on the B cell surface immunoglobulin of the malignant B cells and the lectin DC-SIGN expressed by unusual cells in the follicle. This study also included new methods for spatial analysis of tumor organization and for relating very high dimensional IF imaging to low dimension but larger area examined by conventional multiplex IF used in pathology laboratories. Additional study of tumor organization involved the application of a new technology for spatial transcriptomics. Employing the NanoString GTX device, we were able to identify tumor cells and not immune or parenchymal cells as the hosts for intratumor bacteria. This association of bacteria with tumor cells correlated with a high degree of expression of oncogenic b-catenin pathway genes in the tumor, suggesting that bacterial association with tumor cells contributes to their malignant properties. As part of the development of technology, we contributed to a community effort to specify the best practices in multiplex imaging and are engaged in work on ontologies for immune tissues and cells as well as generation of specified staining panels that will enable easier application of iterative multiplex imaging by more laboratories seeking to use these methods in their research efforts. We also reported on the latest developments in optical imaging as applied to the immune system.
View original record on NIH RePORTER →