December 2-4, 2020 | 10:00AM ET | 3:00PM UTC*
*Program is subject to change
Click on the 'Resources' tab to download the ePoster.
Developing a Multiplexed Imaging Assay for Non-Human Primate Tuberculosis Granulomas
Alea Delmastro1,2, Erin McCaffrey2, Joshua Mattila3, and Michael Angelo2
1Stanford University Department of Biomedical Informatics; 2Stanford University Department of Pathology; 3University of Pittsburgh Department of Infectious Diseases and Microbiology
An estimated one-fourth of the world’s population is infected with tuberculosis (TB). TB infection drives the formation of organized immune cell aggregates known as granulomas. It is thought that immune responses in granulomas drive disparate infection outcomes, but histological correlates of protective immunity in TB are not understood. The inability to routinely sample human granulomas as well as the fundamental differences between human and mouse TB presentation have hampered research in this area. Alternatively, macaques are experimentally-tractable non-human primates that mirror human TB. Here, we describe the development of the biomarker assay, which will be used for multiplexed imaging of macaque granulomas. Tissue pathobiology is typically studied with immunohistochemistry (IHC) approaches; however, these approaches are limited to analysis of 1-3 proteins at once, preventing acquisition of high-dimensional single cell and spatial information. Multiplexed ion beam imaging (MIBI) overcomes this limitation by simultaneously imaging over 40 antibodies in formalin-fixed paraffin-embedded tissues. Our pipeline combines MIBI, deep-learning based segmentation, and single cell analyses to elucidate the composition and structure of granulomas and describe factors that differentiate granulomas within and between individual macaques. We validated >90 reagents via chromogenic IHC for cross-reactivity in macaque tissues with >60 targets optimized for MIBI. Using IHC analysis, we see suspected Mycobacterium tuberculosis antigen secretion in syncytin positive giant cells. In preliminary MIBI imaging, we also observed IL33 expression in the myeloid core. In a previous study, we validated our MIBI workflow and obtained proof-of-concept preliminary results in rhesus macaque splenic granulomas. Our upcoming study contains a cohort of cynomolgus macaque lung granulomas across various stages of TB progression in order to describe key immune differences across the disease spectrum. We will apply the FlowSOM clustering algorithm to identify cell lineages and quantify expression of key functional markers, such as PD-L1, CD40, and phospho-S6, across populations to, for example, quantify T cell infiltration and activation. We will assess the single cell structure of granulomas as it relates to bacterial load, infection time, and the inflammatory profile of granulomas. The results from this study will enable deep characterization of the heterogeneous immune response to TB and may identify immune pathways that could be critical to targeted therapy.